S1 Rectangular Plazas, New
12CPL Constantinople (Plaza by the Church of Forty Martyrs): This plaza consisted of a paved square covering a cistern, built by Phocas in AD 609 with his column: Chronicon Paschale Olympiad 347 ‘AD 609’; Patria 2.64. See discussion of Column of Phocas in appendix F9. This can probably be equated with the cistern described by Mamboury found at this spot. The internal dimensions of the cistern measure 16.6 m by 23.95 m, which must be the minimum dimensions of the plaza, which could have been ca. 40 m by 55 m without disturbing hypothetical shops and porticoes at the side of the Mese (measuring off E. Mamboury, “La nouvelle citerne byzantine de Tchifté Sérail”, Byzantion 11 (1936) 167–80 pl. 17).
Dating summary: 609, class x (historical), publication 3/3.
12CPL Constantinople, Anaplus: An agora was built by the sea by Justinian, according to Procop. Aed. 1.8.6–11 and 17–18, as part of works associated with a church. It is not clear whether the agora is a plaza or merely a market. The context of the work makes it sound like a monumental area. It joins the sea with an ἐµβολὴ τῶν πετρῶν (translated as a ‘stone quay’ by the edition of Dewing (1954), but should perhaps be translated as a colonnade). The dating, if Procopius is correct, should be within the reign of Justinian but before the collapse of the dome of Hagia Sophia in 558. There is no acknowledgement of this in Procopius’ work, despite a description being given of the monument in 1.1.66–78, which discusses problems of its stability.
Dating summary: range 527–58, midpoint 542.5, class x (historical text), publication 3/3.
13ASI Laodicea ad Lycum (Central Agora): This plaza, located on the south side of Syria Street, at the centre of the city is a rectangular plaza, bounded to the north by Syria Street (where there is no portico) and to the south of the North Baths-Gymnasium complex. It is described by its excavator in the following report: C. Şimşek, Laodikeia (Laodicea ad Lycum) (Laodikeia Çalişmalari 2) (Istanbul 2013) 178–83 with figs. 233–238, including reconstruction drawing fig. 235 and pp. 274–96, describing the building of a church in the plaza with a multi-phase plan on p. 276 fig. 371. The ‘central agora’ was substantially rebuilt in the very late 5th or 6th c., but seems likely to have inherited its form from an earlier space, perhaps the forecourt of the baths. An early publication describes the square as a ‘Byzantine’ marketplace: C. Şimşek, “2004 Yılı Laodikeia Antik Kenti Kazısı”, KST 27.1 (2006) (419–34) 427. See also C. Şimşek, “2003 Yılı Laodikeia Antik Kenti Kazısı”, KST 26.1 (2005) (305–20) 307–308; Id., “2005 Yılı Laodikeia Kazısı Çalışmaları”, KST 28.1 (2007) (455–78); Id., The Ancient City of Laodicea (Denizli 2010) 10.
Of architectural form, it is a rectangular plaza, open onto the main east-west street on the north side. It is surrounded on three sides (east, west, and south) by a portico fronted by two steps. The south portico is blank [i.e. with no doorways] against the baths, with the east and west porticoes having doors to shops, according to C. Şimşek, The Ancient City of Laodicea (Denizli 2010) 10. On a photo in Şimşek (2013) 190 fig. 248, the north side seems to have a low wall, likely a stylobate, separating it from the east-west street, but this is only shown as stairs leading down to the road on the reconstruction drawing on p. 181 fig. 235. The agora paving is shown as having a height slightly lower than that of the portico paving of Syria Street leading to the west on this image, with the portico of the agora and the portico of the street joining up, but with ?3 steps needed to reach the agora floor and 3 steps needed from the agora to reach the roadway paving level. In early imperial times there were two portals from the central agora to the baths. One is at the south-west, which was closed later. The other is at the west, which is a two portalled gate made of reused marble and which was also used in a later period: p. 183.
Of dimensions, it measures 112 m by 60 m: Şimşek (2010) 10. The adjacent baths, here marked in black, are 89.6 m by 57.6 m, according also to a note on p. 10. The nymphaeum across the road measures 41.6 m by 14.3 m, according to a note on p. 9. I have tried to reconcile these measurements using the plan in Şimşek (2006) 429, fig. 2, which confirms the accuracy of these figures, and that the agora dimensions are taken from the back wall of the portico.
Of paving, the reconstruction drawing of Şimşek (2013) on p. 181 fig. 235 shows the agora as being covered in large square slabs of around 4 m by 4 m. It is not clear if these are simply a device of the reconstruction artist, as they are not referred to in the text, as far as I could see. Şimşek (2007) 458–59 describes a sondage in which was discovered an earlier Roman paving underneath a ‘Byzantine’ mortar layer, which suggests that at least a mortar layer was present on the agora, for a late paving, unless the sondage was within the portico.
Of porticoes, Şimşek (2013) 179 gives them as 6 m in the south, 5.75 m east and 6.25 m in the west. The horizontal beams of the east and west porticoes stood on the shop wall. That of the south portico horizontal beam was based on the central baths. In the bath wall one can see two huge rectangule-shaped holes and between them are four small ones [from my own visual inspection of p. 180 fig. 234 I make it three large holes with 4 or 5 holes between and small holes of the same type also visible under the large holes]. The horizontal beams carrying the roof were placed on 3.82 m high columns. The height of the floor of the other portico to the horizontal beam is 6.75 m, equivalent to the total colonnade height, perhaps inclusive of architrave: p. 181. The portico stylobate is made by reused marble and travertine. The west and east portico columns were raised up on reused Roman period Attic-Ionic bases [kaideler—my interpretation of the reconstruction drawing is that they are mainly bases, although a few pedestals are shown]. L-shaped plinths were placed at the south-west and south-east corners of the porticoes. Reused Corinthian capitals were found: p. 181. At the south-east part of the agora Attic-Ionic pedestals and ‘Early Byzantine’ impost capitals have been found: Şimşek (2013) 181–83. No arcading is shown on the reconstruction of p. 181 fig. 235.
Of portico paving, the west portico was covered in opus sectile using areas of marble squares and rectangles as well as octagons and circles: p. 181. This is illustrated on Şimşek (2013) 182 fig. 236, where white and black marble were used, in an inconsistent manner, with white predominating. There are three schemes visible as loose bands of ornamentation.
Of internal features, in the centre of the agora was a Roman monument that was apparently converted into an honorific column, perhaps for a cross monument. This is a square structure, visible on the plan and on photos of Şimşek (2010) 10, where it is clearly a monumental structure in ashlar blocks. It seems to be 7.5 m by 7.5 m in size: Şimşek (2013) 183. Its extant remains consist of a square platform with 3 surrounding stairs, built of marble blocks. At the centre of the platform was found a geison sima block (of the Severan period) set upside down, on which was placed a column. Around the platform they found a Corinthian column capital of the ‘Early Byzantine period’: Şimşek (2013) 183. The new-cut Byzantine Corinthian capital fragment is shown on Şimşek (2013) 182 fig. 236. Apparently, the three-step block and a statue head of a woman found here shows that there was originally a Roman monument here: Şimşek (2013) 183. A basin also appears towards the centre of the south side of the plaza, on the reconstruction drawing on Şimşek (2013) 181 fig. 235. On p. 183 it is described as a rectangle-shaped pool and fountain of reused material, with a basin measuring 3.15 m by 1.4 m. It is described as being ‘Early Byzantine’ in date.
Of phasing, all reports also seem to assume this ‘Byzantine’ space is a rebuilding of a Roman one, and the sondage down to Roman paving, mentioned above, confirms this. However, the porticoes including their stylobates are visibly late antique, as is the style of opus sectile, and apparently the paving is as well. It is, however, not clear what kind of space this was, perhaps a forecourt to the baths, of perhaps part of a gymnasium. Thus, I treat this site as a new agora, given the large scale of new construction work and the low level of visiblity of earlier structures. The fountain basin seems, from the reconstruction drawing, to overlap the south portico, so I take it as belonging to the same date range as the agora, set between two earthquakes, rather than being earlier.
Of dating, in all publications that I have seen so far, Şimşek believes in a catch-all dating scenario of repair after a damaging earthquake in AD 494, which is probably what causes him (in the publications above) to call the agora Byzantine, as the present condition, with its reused columns and opus sectile, is likely to represent the post-494 phase on the site. In Şimşek (2010) 10 the adjacent baths are also said to have been rebuilt after the earthquake of AD 494. The earthquake is that recorded as destroying this city and others in the neighbourhood by Marcellinus Comes in AD 494 [Guidoboni Catologo no. 163]. For a discussion of the earthquakes of AD 494 and of 602–610 at Laodicea, see the entry on Syria Street in appendix C3.
Of use, they found some plaques and crosses on the portico stairs: Şimşek (2013) 181.
Dating summary (rebuilding of the agora): range 494–610, midpoint 552, class Cs2 (catch-all, earthquake), publication 2/3 (as earthquake damage is not substantiated).
Dating summary (fountain): range 494–610, midpoint 552, range 494–610, midpoint 552, class Cs2 (catch-all, earthquake), publication 2/3 (as earthquake damage is not substantiated).
Dating summary (honorific column): range 494–610, midpoint 552, class Cs2 (catch-all, earthquake), publication 2/3 (as earthquake damage is not substantiated).
15ORI Bostra (Bosra): A new rectangular plaza has recently been discovered in the centre of Bostra, just east of the tetrakionion, on the north side of the main east-west colonnaded street. The plaza has been excavated on its southern side, revealing a row of shops fronted by a portico, with a triple-arched monumental entrance providing access from the east-west colonnaded street, which was at a higher level than the plaza: C. Delplace, “Le cryptoportique”, in Bosra aux portes de l’Arabie, edd. J. Dentzer-Feydy, M. Vallerin and M. Fournet (Guides Archéologiques de l’Institut Français du Proche-Orient 5) (Amman 2007a) 257–60 with an outline plan on p. 260; id., “Le forum”, in the same book (2007b) 261–62. For early reports, see J.-M. Dentzer and P.M. Blanc, “Le centre de la ville de Bosra: travaux de la mission en 1998”, AAS 44 (2001) (131–52) 132–34 and especially C. Delplace and T. Fournet, “Bosra 1998 - nouvelles recherches dans le secteur du cryptoportique”, AAS 44 (2001) 153–65; C. Delplace and T. Fournet, “Bosra 1999. Deuxième campagne de fouille dans le secteur du cryptoportique”, AAS 45–46 (2002–2003) 305–16.
Of dimensions, no overall detailed plan of the plaza is available, so its size cannot be fully estimated. Only one corner has been excavated. However, from the photographs (Delplace (2007b) photo on p. 262, plan on p. 260), an area of around 50 m by 25 m seems to have been excavated. The maximum possible area for the plaza before it hit the shops lining the street to the north or east, or the macellum to the west, or a known wall slightly further north, is 91 m to 98 m east-west by 81 m to 106 m north-south, hand-measuring roughly off the plan. The space created would thus be slightly trapezoidal. The maximum extent proposed here is likely to be correct, as it would put the only known entrance into the plaza at the centre of the south side of the plaza.
Of the monumental entrance, which was cut through the portico into the plaza, we have a plan in Delplace and Fournet (2001) 155 fig. 3. The same report informs us on p. 162 that the central portal of the entrance measures 3.4 m wide and was flanked by two side portals of 2.4 m wide. All portals were likely covered by arches. On the plan fig. 3 it is clear that the walls of this monumental entrance, although essentially just ‘a wall with entrances’, are wider than the wall of the pre-existing shops, and seem to be secondary to them. The façade was set to correspond with the back wall of the shops, and the excavators suspect that it was fronted by columns set to correspond with the position of the doorways of the shops, to create a monumental vestibule, an ornament that would have been appropriate either on the face onto the street or that onto the plaza. However, this is speculation, as the walls that would have carried them do not seem to have survived, from the plan.
Of porticoes, only the south portico is known, and this is ca. 6 m wide, from back wall to front of stylobate, hand-measuring off plan fig. 3. The portico has been excavated only in short stretches. A photo on p. 262 also appears to show the presence of a stone slab paving within the portico.
Of paving, the rectangular slabs were laid in rough rows of irregular width, sometimes interrupted when larger slabs were used, as shown on p. 262. The presence of some slabs of very variable sizes in front of the south portico makes it possible that some of them were reused, although there are no visible examples of angles being cut into slabs to make different slabs fit. By combining the plan on fig. 3 with the photo, I was able to roughly estimate that the size of the slabs varied between ca. 0.4 m square and ca. 1 m square, though there are many rectangular slabs of dimensions between these two extremes.
Of the colonnades, columns, capitals and hexagonal bases have been found relating to the portico, see: Delplace (2007b) 261 with a drawing of a capital, which is clearly Ionic. There are no details provided of the size of the pieces or the materials they were carved in.
Of shops, 11 one-room cellular units have been excavated, without knowing if any are interconnected, or revealing more than a couple of thresholds. The depths of the shops are all ca. 4.3 m on the east side, but those on the west side are varied, with two or more being 4.8 m and one being ca. 2.3 m. Given that the plan of their walls, Delplace and Fournet (2001) 155 fig. 3, shows many uncertainties, it is not really sure if these shops are late antique or not. They vary in width from 2.3 to 5.6 m, not counting one room cut by the monumental entrance wall. All these measurements are derived from hand-measuring off the plan of fig. 3.
Of phasing and dating, the first recorded element on site is an Early Imperial cryptoporticus that supported a portico and a row of (south-facing) shops that lined the east-west colonnaded street. Two features of the agora were secondary to this and seem to be late antique. Firstly, the triple-arched monumental entrance was cut through the cryptoporticus shops from the east-west street (Delplace (2007a) 258 with map on p. 260). Secondly, the plaza shops and portico (north-facing) were built against the back of the cryptoporticus shops, i.e. against their north side, a detail I was not able to confirm from plan 3, which did not seem to show this. We also have stratigraphic studies of Delplace and Fournet (2001), for which see especially pp. 161–65. For these authors, writing in 2001, the plaza and its monumental entrance belong in phase 2, and the installation of a ramp in the cryptoporticus to phase 3. Both phases occur before the round plaza (of the 4th c.) in phase 4, a statement that seems to correlate with the walling phases of both structures as they are shown on Delplace and Fournet (2001) 155 fig. 3. This understanding seems to have been modified by the time of the 2007 publications, in which (on p. 258) the ramps, replacing earlier staircases, are stated as being 4th c., being installed when the cryptoporticus floor was raised by 60 cm, a judgement that likely depends on finds (p. 258). The plaza paving is now considered 5th c. (p. 261), based on finds from sondages beneath the agora. These sondages are described in most detail in Delplace and Fournet (2002–2003) 310–12. In fill layer 4075 (heavy red earth) was found “céramique rouge locale, grey ware byzantine, terre sigillée claire, céramique à pâte claire à structure feuilletée à engobe bleuté, et fragments de céramique nabatéene”, which suggested to the excavators a date of towards the 5th c. [I have not been able to identify specific dated wares from the names provided], whilst brown fill layer 4078 was found to include “céramique à pâte rouge des Ier–IIe s., à pâte beige du IVe s.”, which suggested a date towards the 4th c. (pp. 310–11). A final sondage under the paving but close to ramp 2 revealed dump layers (4091–4098, 4099) dating from the 5th c., containing Cypriot ceramics of the 4th–5th c., an amphora of an unspecified Riley type of the end of the 5th c., and Aegean amphorae with a micaceous fabric, imitations of African pottery plus red and grey fabric wares (p. 312). These layers suggested a date of the end of the 5th c. for the plaza to the excavators (p. 312). However, I will take the date as being the 25 years centred on AD 500, assuming the Riley amphora was the latest datable find, as it dates from the end of the 5th c. (AD 487.5–500), which under the rules of this study generates a contextual date of 487.5 to 512.5.
Of subsequent phasing, layers excavated immediately over the paving (reflecting occupation) contained unspecified pottery of the end of the 5th to 6th c., whilst brown soil covering the paving (representing degradation) contained 7th c. material (which seems to be ARS, Jerash bowl, sigillata of Umayyad date, Gaza Ware and Grey Ware etc. [from which names again I was not able to do a successful external check on the dates]). The full publication of the ceramics, with more specific naming of wares, is of course needed, except perhaps for the last deposit described here, which has a fuller description. The brown earth over the paving was covered in part with a major stone destruction layer (p. 310). I take the brown soil as representing serious degradation, with the previous layers representing occupation, even if the loss of parts of the paving prior to that, under occupation layers, represents a degree of downgrading.
Some doubts remain about the phasing. It is not certain that the new entrance was established at the same time as the extant plaza paving, although this is likely. Delplace (2007b) 261 expresses some doubts that the plaza was established entirely anew in the 5th c., noting that the excavators have not yet identified any earlier structure within the architectural frame provided by the surrounding buildings. However, the ‘lost’ structure might well have been a former temple temenos or some other open space: it need not have been an agora. The fact remains that we have a paved square with monumental entrance for which there are no indications of a date before the 5th c., within a prosperous city with much evidence of late antique building. Thus, the dating, although poorly published, within phasing that seems still to be provisional, must be allowed to stand, using the full range offered for the ceramics in the report.
Dating summary (construction of plaza): range 487.5–512.5, midpoint 500, class Cs9 (contextual ceramics), Cs4 (reused material), publication 2/3.
Dating summary (covering of paving by occupation layers): range 487.5–600, midpoint 543.75, class Cs9 (contextual ceramics), publication 2/3.
Dating summary (serious degradation of plaza): range 600–700, midpoint 650, class Cs9 (contextual ceramics), publication 2/3.
16AEG Abu Mina (Pilgrims’ Court): The Pilgrims’ Court was the main square of Abu Mina. It served pilgrims coming towards the church, acting as a climax to the colonnaded street, before they entered the sanctuary complex. It had a central monument / honorific column, which was aligned so as to be visible when coming into the square from the colonnaded street. However, the square does not serve the church itself in providing any kind of a direct approach to the building, as might an atrium. It does, however, provide a plaza for people coming out of the church. For reports see P. Grossmann et al., “Abū Mīna Zehnter vorläufiger Bericht. Kampagnen 1980 and 1981”, MDAIK 38 (1982) (131–154) 141; P. Grossmann et al., “Elfter vorlaüfiger Bericht. Kampagnen 1982 and 1983”, MDIK 40 (1984) (123–51) 139, with overall plan on p. 135, and detailed plans (stone plans and phase plans) of the north-west corner on pp. 140–41 figs. 5–6.
Of architectural form, the plaza was rectangular. It is particularly interesting because it is an example of a rectangular plaza that replaces an elliptical open space, not a rectangle reusing an earlier quadrilateral plot. Its design demonstrates that a rectangle was the preferred choice of shape, not one imposed by circumstances. On the north side, the axis of the north-south colonnaded street (here not colonnaded) enters through a wall-arch described in appendix F8a. It has a colonnaded portico on all 4 sides. On the south side, the portico fronts the Tomb Church and Great Basilica, whilst on the other sides the porticoes front courtyard buildings (probably xenodocheia) and some cellular rooms, which may be shops.
Of dimensions, the plaza measures ca. 35 m by 85.8 m (from back wall to back wall) according to my hand-measurements off the overall plan, with the north portico closer to 85 m, being slightly shorter than the south portico of 85.8 m.
Of porticoes, the north portico (towards its west end) measures ca. 3.15 m, from back wall to front of stylobate. The west portico measures ca. 3.35 m (towards its north end), hand-measuring off the plan on: Grossmann et al. (1998) 140 fig. 5. The east portico seems to measure between ca. 3 and 3.5 m, hand-measuring off the overall plan. The south portico is the most varied in width, as it tries to mask the uneven edge of the Tomb Church and Great Basilica, measuring ca. 2 m to 3.35 m, hand-measuring off the overall plan. The remains of the colonnade in the north-west corner of the plaza give an average interaxial measurement of 2.23 m on the north and 2.42 m on the western colonnade: Grossmann et al. (1984) 139. The portico outside the Tomb Church is thought to have first had a sandstone paving, 8 cm thick: P. Grossmann, Abu Mina, vol. 1: Die Gruftkirche und die Gruft (Archäologische Veröffentlichungen 44) (Mainz 1989) 62. This was replaced by a paving of stone slabs, in association with the Justinianic church. These slabs were in marble around the church entrance: Grossmann (1989) 63.
Of the colonnades, work in the north-west corner of the Pilgrim’s Court, mainly on the later stratigraphy, notes that pedestals of the colonnade, in marble and nummulite limestone, are preserved in situ: 9 on the northern colonnade and 3 on the western colonnade (plate 7a). No complete column shafts survive, although a reconstructed example suggests they were ca. 2.1 m high. The absence of architraves suggests that the capitals might have held arcades rather than flat architraves. The excavators suggest this, noting that all the other walls of the porticoes were built only of sand-limestone. In the corner of the plaza was a T-shaped pier rather than a pedestal: Grossmann et al. (1984) 139. In the western portico, a kind of cornice (perhaps a bracket for a tie beam or to support the entablature of the roof) and one console block were found, both of local limestone, and believed to be part of the portico: p. 142 with n. 64 and 145 fig. 8. A suggested reconstruction of the portico colonnade is on p. 146 fig. 9. In the south portico, outside the Tomb Church, a pedestal with conjointed column bases has been found in situ on the western part of the colonnade. It is ca. 1.64 m high and supports a column of ca. 0.38 m diameter at its base, if my very rough measurements off a photocopy of Grossmann (1989) 61, fig. 9 can be relied upon.
Of paving, the square seems to have been covered by what looks like new-cut stone slabs sorted into straight rows of uneven width, arranged in more than one system of alignment: see the map on Grossmann et al. (1984) 141 fig. 6, and p. 140 fig. 5 for the paving of the plaza. Slab sizes vary from ca. 0.42 m by 0.48 m to ca. 2.4 m by 1.1 m, hand-measuring off p. 140 fig. 5.
Of shops, the outline of cellular units set along the north side of the square is known, but there is nothing except a slight acknowledgment in the report text to suggest that they belong to the same phase as the portico: Grossmann et al. (1984) 139. They are well-planned on p. 140 fig. 5 and p. 141 fig. 6, revealing 6 units of irregular size, all opening onto the plaza, of which 5 are one room and one has a second backroom extending off it to the rear. They are rectangular, except for the westernmost, which tries to adapt to an irregular back wall inherited from an earlier phase, and so is triangular: Grossmann et al. (2006) 39 fig. 4. Their internal dimensions, excluding the triangular unit, are of irregular sizes, measuring ca. 1.9–3.25–4.4 m deep and ca. 1.8 m to 4.2 m wide, hand-measuring off the plan. Other individual rooms around the plaza may have served as shops, but these are often connected to the structures behind in some way or have not been fully excavated.
Of monumental furnishings, the square has a fountain, set almost at its centre, facing down the axis of the north-south colonnaded street as it joined the square: Grossmann et al. (1982) 141. No plan is given, but plate 20c is a photo that shows one side of the fountain, its perimeter drain and a drain leading away from it. On p. 141 it says that it was supplied with fresh water, and that lead pipes were found around the fountain. The shape of the building implies that it was a 4-piered building [so like a canopy]. In Grossmann (1984) 138, the author calls the fountain in the centre of the Pilgrim’s Court a ‘tetrapylon’. The square also features a foundation of what may have been an honorific column, set at the exact centre of the plaza, for which see appendix F9.
Of phasing and dating, the main sequence for the Pilgrim’s Court square comes from the south portico and its relationship to the churches behind: this is best studied outside the Tomb Church (Gruftkirche). It is a highly complex sequence, but must be understood to grasp the chronology of the square and of the whole settlement, as relatively little stratigraphic excavation has taken place here, and so few finds have been recovered.
Firstly, the Tomb Church (when it was still a basilica) was given a north portico. This happened at the same time as the church was given a baptistery: we can say this because the portico extends to incorporate the baptistery as well as the church, and also because the portico displays an interaxial spacing that shows that it pre-dates the subsequent Justinianic church with its different entrance: Grossmann (1989) 60–63. The baptistery phase is not dated, but must date slightly later than the 5th c. basilical phase of the Tomb Church, but before the Justinianic phase. The basilica phase is dated to the first half of the 5th c. based on pottery (not specified here, but perhaps the context of ca. 475 below): Grossmann (1989) 14, 38–50, plan 3.4–5.
Secondly, the portico in front of the Tomb Church was modified in its eastern section, when its rear wall was masked by the construction of the Great Basilica to the east in the late 5th c., with plan of J. McKenzie, The Architecture of Alexandria and Egypt 300 BC–AD 700 (New Haven, Connecticut-London 2007) 289, fig. 481. This church had two major phases, the second of which dated to after ca. AD 475. We know this because a cistern next to the first apse of the church, sealed by walls of the second phase, contained a fill with pottery that was developed by this time [the use level and the back-fill of the cistern showed no chronological differences]: H. Schläger, “Abu Mena. 2. Vorläufiger Bericht”, MDIK 19 (1963) (122–25) 123 with fig. 1 and plate 41a (showing the relationship of the cistern and the walls), as pointed out by McKenzie (2007) 291, 417 n. 148, who notes its significance in providing a TPQ of phase 2. The pottery is described by J. Engemann, “Kleinfunde und Keramik”, in W. Muller-Wiener, “Abu Mena. 4. Vorläufiger Bericht”, MDIK 21 (1966) 185–87, and the key ceramics are LRC (Phocaean Red Slip Ware), which J. Hayes has assured the author had reached the whole east Mediterranean by the end of the third quarter of the 5th c., alongside LRB-E, with D-E being especially numerous. The absence of obviously later material suggests a date of this time or shortly afterwards for the end of the formation of the deposit.
Thirdly, the portico in front of the Tomb Church was then later extended eastward to cover the entire north side of the Great Basilica (being set at its height), forming now the south portico of a wider rectangular plaza of which it was a constituent part: Grosssman (1989) 60. N.B. the plan above shows that the portico in front of the Great Basilica is aligned to the rest of the square, but not with the church, and so belongs with the Pilgrim’s Court square as a whole. The fountain and the monument base in the centre of the plaza probably belong to this phase, given their careful setting in and alignment with features of this phase.
Fourthly, around the same time, or perhaps a bit later, the western part of the portico in front of the Tomb Church must have been altered, as the church was rebuilt under Justinian, involving the reconstruction of the rear wall of the portico: Grossmann (1989) 137–78. This rebuilding of the church is dated not before AD 532 based on unspecified pottery: Grossmann (1989) 15 with n. 13 (who mentions the pottery is being looked at by J. Engemann, though I could not locate a subsequent publication), 97. The precise date of ‘after 532’ for the ceramics seems suspicious, and might relate rather to the foundation date of the Justinianic Hagia Sophia for some reason. To access the new Tomb Church, a projecting porch with piers was added, interrupting the western part of the portico, which seems to have involved replanning some of the colonnade spacing in this section: Grossmann (1989) 137, but not projecting from it. A detailed stone plan of the area is available in W. Müller-Wiener, “Abu Mena 3. Vorläufiger Bericht”, MDIK 20 (1964) 126–37, fig. 3 which shows the two square column / pedestal bases within the portico, nearest to the door, which they frame. If we take a pedimental porch (of which the upper architectural parts are not confirmed) as being set onto these two bases then the porch was ca. 5.8 m wide, roughly hand-measuring off the plan, and as wide as the portico, which here is ca. 3.5 m wide, from back wall to front of stylobate. Subsequently, the portico just east of here was covered in part by a tribunal: Grossmann (1989) 139–42. All these phases date to before the extensive fire that is supposed to have occurred in the ‘early 7th c.’ (Persian sack of 619 theory, see appendix C3), which left traces of burning in the portico, on the tribunal outside the north wall of the Tomb Church: Grossmann (1989) 142.
Of phasing and dating, other areas also give small clues. On the north-east corner of the square, the porticoes are clearly secondary to walls that run underneath, and here have been given a tentative chronology of the early 6th c., although it is not stated why: perhaps it is based on the relationship of the outer portico to the Justinianic rebuild of the Grave Church described above, though we have no published grounds to accept this here: J. Kosciuk, “6: the north-west corner of the Pilgrim Court”, (139–47) 139, in P. Grossmann et al., “Elfter vorlaüfiger Bericht. Kampagnen 1982 and 1983”, MDIK 40 (1984) (123–51) 139. Recently a large hemicycle, predating the great rectangular square of the Pilgrim’s Court, has been detected from sondages that can be linked with adjacent buildings, and which was replaced by the rectangular Pilgrim’s Court and its surrounding xenodocheia: P. Grossmann and J. Kosciuk, “Report on the Excavations at Abu Mina 2005”, Bulletin de la Société d’Archéologie Copte 44 (2005) (29–44) 42–43 with p. 39 fig. 4. This hemicycle seems not to have been connected to the colonnaded street or to the late 5th c. Great Basilica in its layout, so presumably predates them both. It had a north-south width from the back wall of the portico to the flank of the Tomb Church of 36 m (hand-measured off p. 39 fig. 4). Its east-west length could be estimated as ca. 67 m based on its curve being identical to that of the southern hemicycle, which it seems to echo in its plan. I obtained the second dimension by hand-measuring the southern hemicycle curve off the overall site plan.
Overall, in the absence of any further published indications, the rectangular Pilgrims’ Court square must be dated using the sequence from the Tomb Church and Basilica Church on its south side. Here it is clear that the portico phase relating to the rectangular plaza dates to after ca. 475 but before AD 532, after which time some rebuilding of the western end of the southern portico took place. It is not clear when the earlier hemicycle was constructed, though it must have been prior to ca. 475. It seems to have been similar in size to the southern hemicycle (p. 32 n. 26), south of the Tomb Church, and indeed seems to frame it and its baptistery. Given that there is apparently no evidence of occupation on the site before the late 4th c. (perhaps based on ceramics?): P. Grossmann, “The pilgrimage center of Abû Mînâ”, in Pilgrimage and Holy Space in Late Antique Egypt, ed. D. Frankfurter (Leiden 1998) (281–302) 282. It seems reasonable to place the hemicycle north of the church within the period 400–75, constructed sometime after the church itself. The use of the cistern deposit to provide a date of 475 for its end of use does seem reasonable, as the deposit is likely related to the use of the cistern rather than being a fill dumped in from elsewhere. The date of 532 is obviously provisional, pending full publication and explanation of its context. The last phase can be given a TAQ of the Persian ‘sack’ of 619.
Dating summary (hemicycle Pilgrim’s Court): range 400–75, midpoint 437,5, class Cs3 (associative, phase of development), Cs9 (contextual ceramics), publication 2/3.
Dating summary (rectangular Pilgrim’s Court): range 475–532, midpoint 503.5, class Cs3 (associative, phase of development), Cs9 (contextual ceramics), publication 2/3.
Dating summary (rebuilding of portico by the Tomb Church, with associated porch): range 532–619, midpoint 575.5, class Cs9 (contextual ceramics?), Cs3 (associative, phase of development), Cs2 (catch-all, Persian sack), publication 2/3.
Dating summary (building of tribunal by the Tomb Church): range 532–619, midpoint 575.5, Cs9 (contextual ceramics?), Cs3 (associative, phase of development), Cs2 (catch-all, Persian sack), publication 2/3.
Discounted: I have not included the secondary hemicycle and rear court of Abu Mina, as unlike the Pilgrims’ Court they are internal plazas of the ecclesiastical complex, which are not integrated within the street system.
S2 Round Plazas, New
09DAC Justiniana Prima: V. Kondić and V. Popović, Caričin Grad: Utvrđeno naselje u vizantijskom Iliriku (Belgrade 1977) 51–54 and 322–23 with p. 50 fig. 30 (plan) and p. 51 fig. 31 (sections). The plaza was levelled despite an inclination, so in some places the portico was higher than the square and in other places it was lower.
Of dimensions, the plaza measures ca. 22 m across, according to the text, but the plan shows this is the open area of the plaza only. Including the porticoes, it measures ca. 30 m from back wall to back wall.
Of porticoes, the width measures ca. 3 m, according to the text, and ca. 4 m on the plan from back wall to front of stylobate and are divided into 4 sections / quadrants by the crossroads coming into the plaza. The stylobate of the portico only survives in part, as is clear from the dotted lines on the site plan, but the text reports that the colonnade is composed of piers at each end of the quadrant with two columns between them.
Of paving, the plaza was covered with stone slabs, more or less the same size, of no more than 1 m2. The plan shows that the slabs covering the plaza are irregular in form, not especially rectangular, and that the porticoes were paved with the same material, in contrast to the adjacent colonnaded streets, where diagonally-laid brick tiles can be found (see appendix C2). A drain ran around the edge of the plaza: p. 322.
Of possible arches (4), over the streets leading into the plaza, we have some suggestive structural evidence. The projection of the end-piers of each semicircular portico, at its junction with an axial road, suggested to the excavators the existence of vaulted entrances into the plaza (p. 322 and also visible on the plan). The plan reveals that not all of the piers survive. Yet, where they, do this thickening and projection of piers is visible in both the piers on the plaza stylobate and in piers immediately behind them. The second set of piers aligns with the back wall of the plaza portico. They are visible on both sides of the road for the west street (on the piers that coincide with the plaza portico back wall); on the west side of the road only for the south street (on the piers that coincide with the plaza portico back wall and the stylobate); on the north street (on the pier on the east side of the road that coincides with the plaza portico stylobate). Based on this evidence, it seems likely that the plaza had 4 (the east is conjectural) vaulted entrances over each of the entrances into the plaza, although this might have been with a wooden superstructure rather than masonry, as the piers are only a little bit wider than the other piers of the street porticoes, only in one case being twice as large. Thus, they do not compare to piers of tetrapyla for example.
Of internal features, a stone ‘column-support’ (“support de colonne”) stood at the centre of the plaza, which may have held the bronze statue (for an emperor given the military dress) of which two fragments were found nearby: p. 322 and p. 33 fig. 33. No direct indication is given of this being an honorific column rather than a columnar statue base, so I prefer not to assume we are dealing with an honorific column here. Shops are not found on this plaza, although an odd door on the south-west side suggests they were planned, along with others on the colonnaded street going west from the plaza, where three equally-spaced doorways have no corresponding rooms behind.
Of dating, this plaza was a critical organising element of the site, as the junction of its axial roads, with which its walls seem to form a single phase. Whether or not this site was originally intended as the intersection by the first architect, it was built as a key part of the initial realisation of the city beyond the acropolis, with its surrounding roads seeing substantial levelling activities. Dating is based on overall occupation evidence from the first phase of this new town, and its supposed identification as Justiniana Prima. We can note that this city, often identified with the city founded near Justinian’s birthplace of Tauresium by Nov. 11 in AD 535 (with Procop. Aed. 4.1.19), has, according to recent researchers, produced no evidence for occupation before ca. AD 530 and no evidence after AD 615 [616 being the date of decisive Slav invasions of this area and definitive loss of Roman control], on which see appendix C2.
Dating summary: range 535–616, midpoint 575.5, class x (historical), Cs3 (associative finds), z (regional background), publication 3/3.
10MAC Dyrrachium: This round plaza, surrounded by cellular rooms, was found at the centre of the fortified city: J. Wilkes, B. Shkodra, E. Metalla, and A. Hoti, “The early Byzantine circular forum in Dyrrachium (Durrës, Albania) in 2002 and 2004–2005: recent recording and excavation”, Annual of the British School at Athens 103 (2008) 367–97, esp. pp. 371–79, with plan on 370 fig. 2. See also A. Hoti, E. Metalla, and E. Shehi, “Recentissimi scavi archeologici a Durazzo 2001–2003”, Antichità Altoadriatiche 54 (2004) 401–33 (not seen); A. Hoti et al., “Scavi recenti 2001–2003 e scavi recentissimi 2004–2007”, in Dyrrachium II: Scavi d’emergenza nella città di Durazzo 2001–2007, edd. S. Santoro and A. Hoti et al. (Athens 2010) (not seen).
Of dimensions, the plaza measures some 52.6 m in diameter, counting from the back wall of each portico, or ca. 40 m counting from the front of their stylobates (hand-measuring off the plan of Wilkes, Shkodra, Metalla and Hoti (2008) 370 fig. 2), east-west across the plaza. Unfortunately, the whole plaza could not yet be excavated, meaning that the main entrance does not seem to have been uncovered. We have rather a round paved space, surrounded by a portico.
Of the portico, the width is ca. 6.3 m on the east-west sides, though ca. 6 m on the north-south sides. There is no information published as to the paving of the portico, but no stone was plotted on the plan, in contrast to the adjacent plaza, making it likely that this surface was in mortar, gravel or a mixture of both.
Of the colonnades, the stylobate of brick-faced concrete held reused limestone plinths (with some inherited features such as anathyrosis and mason’s marks) supporting Proconnesian marble column bases of 0.27–0.29 m high, ‘quarry-ready’ in style (referred to as being like inverted Doric capitals), “marble” column shafts (one surviving intact is 3.45 m high) and new-cut marble Corinthian capitals of 0.57 m high: pp. 371–76. Thus, the colonnade measured ca. 4.29 m to 4.31 m high.
Of paving, the plaza was paved in grey-white rectangular marble slabs, 0.6 m thick and apparently new cut, with no signs of reuse. They covered all of the plaza, except for a strip 5.2 m wide around the central rotunda. The slabs (measuring between ca. 0.5 and 3.4 m in length, hand-measuring off the plan, and similar varied widths) were laid in rows of different widths in a concentric radial arrangement, tapered to fit, except for an outer row of larger slabs, 0.82–0.9 m wide, laid lengthways off the outer edge of the perimeter drain (pp. 376–77 with p. 370 fig. 2). This shallow drain ran immediately in front of the colonnade stylobate, right round the plaza, and exited via a gap in the surrounding rooms, which seems likely to be an entrance to the complex: p. 371.
Of statuary, at the centre of the plaza is a round foundation, called a ‘rotunda’ in the reports, of mortared rubble brick and stone, 5.75 m in diameter, which supported a square stone plinth of 1.82 m by 1.8 m, composed of two massive stone blocks. The core of the foundation reveals that this monument had at least two concentric steps of 0.25 m high. It is estimated by the excavators [likely from its form and width] to have originally extended at least 3 m above the marble paving (p. 379 with p. 370 fig. 2, p. 392 with fig. 13, plates 40 and 45c). There is a strong possibility that this foundation held an imperial statue, likely of Anastasius, as this city was his birthplace, perhaps set on an honorific column (p. 395).
Of shops, the plan shows 8 one-room units, which are oddly trapezoidal in shape, as if twisted by the ‘spinning’ of the plaza, being in 3–4 cases wider at the outside of the complex than the inside (ca. 9 m wide on the outside, ca. 4 m to 5 m wide on the inside), and in 3–4 cases approximately the same width at both outside and inside (ca. 4 m). The shops are also of different depths, but they fit within a space defined by an ‘outer circle’ wall and the ‘inner circle’ of the portico rear wall (a space ca. 7 m in width). These are all approximate hand-measurements from the plan. Although the footings of the shop walls do survive for a few courses, there is no indication that they were entered from the plaza, with three doorways apparently recovered facing outwards onto other spaces. Either the thresholds were higher up and do not survive or were not present. There is no indication, in the report, of what their floors were made of, or what the walls were covered with. From the plan, what may be a staircase seems to be present in one shop.
Of phasing and dating, see especially B. Shkodra-Rrugia, “Late Roman coarse and cooking wares from Durrës”, in LRCW 3: Late Roman Coarse Wares, Cooking Wares and Amphorae in the Mediterranean. Archaeology and Archaeometry. Comparison between Western and Eastern Mediterranean (BAR International Series 2185.II), edd. S. Menchelli, S. Santoro, M. Pasquinucci and G. Guiducci (Oxford 2010) 721–730, complemented by B. Shkodra, “Ceramics from Late Roman contexts in Durrës”, Annual of the British School at Athens 101 (2006) 427–57, which contains rather less detail. The phasing essentially consists of the primary build and a small amount of rebuilding detected in the tabernae around the plaza. Unfortunately, nowhere is this rebuilding mapped in the existing reports, and one is given the impression that it is a rebuilding that has been detected more in ground layers than in the wall fabric, which seems to be part of a single phase development. In contrast, the dismantling of the complex seems much clearer. Shkodra-Rrugia (2010) 721–22 explains the dating best, by examining a number of key ceramic assemblages, from groupings of contexts, which in each group seem to relate to the same events. However, she does not support her chronology by presenting dates normally ascribed to different individual wares. Thus, it is hard to see how the group dates have been arrived at. Furthermore, the methodology used in this study for dating is a little different to that used in the site reports. Thus, the remarks made below represent in many cases my own interpretation, rather than simply repeating the view of the report, although I am undoubtedly at a disadvantage to the excavators, in not having seen the evidence firsthand.
(i) Group 1, from a redeposited fill made to create a walking level, from trench north of the central podium, is given by her as being of the 4th to 5th c. Diagnostic wares are a thin-walled dish of Hayes 50B [given as B—ca. 350–400+ LRP], spatheion Bonifay 1B = Keay 25E [387.5–450 Bon], amphora Robinson M273 [300–500 RADR], plus sherds of amphorae LRA1 [350–650 on RADR, beginning 387.5 for first subtype LRA1a on Pieri 2012], LRA3 [387.5–?600 RADR, for two-handled] and LRA4 [not in RADR]. This evidence, as presently known to me, suggests a deposition date for group 1 of 400–25 based on the start date of the last dated ceramic. However, the fact that the fill has been redeposited means that the date gives us only a rough TPQ for the walking level of the plaza.
(ii) Group 2, from contexts relating to the construction of the plaza, is given as being of the second half of the 5th to the first quarter of the 6th c., from trench north of central podium, providing a TPQ for the construction of the Rotunda. Diagnostic wares are ARS Hayes 61B [ca. 380–450 in LRP with SLRP, but Bon 400–500, which I now follow] and 50B/64 [64 being 375–450 Atlante p. 87 quoting LRP and Michigan I p. 85], East Mediterranean Red Slipped Ware Hayes 3E [could be LRC Hayes 3e of 487.5–500 in Atlante or Cypriot Red Slip Ware Hayes 3 of 475–550 in LRP], African lamp of classic type Atlante X given as of the second half of the 5th c. [Bon 425–700], spatheion type Bonifay 2A [450–500 in Bon esp. p. 127]. This makes the spatheion the latest certain ceramic, which would give the group a deposition date of 450–75, but the East Mediterranean Red Slipped Ware Hayes 3E (likely LRC) suggests a date of 487.5–512.5.
(iii) Group 3, from a well deposit relating to the construction of the plaza (from the south-west side of the monument, near the peripheral wall of the colonnade), is given as being of the second half of the 5th to the first quarter of the 6th c. Diagnostic wares are Hayes 87C [500–512.5 LRP], Phocaean Red Slip Hayes 3F [500–600 Atlante], East Mediterranean amphora types such as LRA1 [350–650 on RADR, beginning 387.5 for first subtype LRA1a on Pieri 2012] and Gaza / Ascalon LRA 4 amphorae [not in RADR] followed by Adamschek RC 22 (of unknown provenance) [not in RADR]. Of this, the Phocaean Red Slip Hayes 3F seems to be the latest piece, suggesting a date for the deposit of 500–525. As this is the latest of the three construction deposits recovered, we can take it as best indicating the construction date for the plaza.
(iv) Group 4, from the tabernae around the edge of the plaza, relates to a reconstruction phase within the 6th c. This reconstruction is not mentioned as affecting the open space of the plaza, the central monument or the portico, which do not seem to have produced any evidence of a second phase. Nonetheless, changes to the tabernae likely affected the roof of the portico, which used it as its back wall. The ceramic finds of group 4 seems to have the same consistency through all deposits. However, the close dating of these contexts is difficult, because two chronologically separate groups of pottery have been detected: there is a dominant group of ARS of the late Vandal period (91C [ca. 530–600+ LRP], 87A–C [C being the latest ware, of 500–512.5 LRP], 99A [487.5–550 Bon] and 99B [525–612.5 Bon]) followed by a lesser number of wares somewhat, dating to after ca. 550 (99C [587.5–612.5 Bon], 109 [c.610/20-c.680–700 SLRP], 105 [587.5–700 Bon], 108 [600–612.5 LRP]). The large number of joins between these contexts caused Shkodra to consider that “the deposition of these contexts is contemporary”, probably implying that they represent use finds from the shops in which they were found. It seems likely that the contexts relating to two different events have been mixed together here, in antiquity or during excavation, only one of which is directly related to the reconstruction of the shops. Here, I attempt to separate them into two sub-events. The first sub-event would seem (from the latest finds) to date from the 25 years after 530, so 530–55, as this is the start date of Hayes 99B, whereas the second sub-event seems to date from 600–25, on the basis of its last ceramic Hayes 108.
(v) Group 5, relating to the dismantling of the complex, was found in the south-west part of the colonnade, lying directly over the pavement. Diagnostic finds included ARS Hayes 104C [550–650 Bon] as well as Phocaean Red Slip Ware Hayes 10 [ca. 570–660+ LRP]. This evidence suggests that the dismantling took place in the period 570–95, although this is a contextual date based on only two wares, which makes it comparatively weak. A large number of graves were documented directly over the marble pavement, with grave goods that are given as being of 7th–8th c. date, with reference given to A. Hoti, “Gërmime arkeologjike të vitit 1987- Durrës (qytet),” Iliria 2 (1987) 261–62 (seen but not able to read). I will not concern myself with the reliability of this last form of dating, as my main interest here is the disuse of the plaza, not its afterlife.
Overall, the following conclusions can be suggested for phasing and dating. These are obviously provisional, as a full contextual publication is not available yet. (i) Firstly, we have a redeposited fill, represented by group 1, made to create a walking level that under the rules of this study seems to have been deposited here in 400–25, based on the start date of the last dated ceramic. However, because it was a redeposited fill (arguably moved from a much earlier primary deposit) it can only provide a rough TPQ for the plaza. (ii) Secondly, we have deposits relating to the construction of the plaza, given by Shkodra as of the second half of the 5th to the first quarter of the 6th c., of which the latest ceramic gives the group a revised contextual deposition date of 487.5–512.5, under the rules of this study. (iii) Thirdly, we have a well deposit relating to the construction of the plaza, represented by group 3, given by Shkodra as being of the second half of the 5th to the first quarter of the 6th c. However, the latest ceramic in this deposit rather suggests a revised contextual date for the construction of the plaza of 500–525. It is this well deposit that seems to provide the best basis for dating the plaza, as it is the latest of all the dating evidence related to this primary phase of construction. (iv) Fourthly, we have Group 4 from the tabernae, relating to a reconstruction phase, which is given as sometime in 500–580 by Shkodra. However, the two separately-dated pottery groups suggest a mixture of two different events, from 530–55 and from 600–25, of which the former might represent the renovation of the shops, whilst the latter might represent its abandonment, considering wider Balkan urban development. (v) Fifthly, we have Group 5, related to the dismantling of the complex “by the 7th c.” according to Shkodra, from the south-west part of the colonnade. From the few ceramics recovered, we can suggest a weak contextual date, for dismantling, from the period 570–95, followed by or contemporary with the development of a cemetery over the area. Given that the ‘abandonment finds’ of 600–25 are close in time to this, we could extend the date to 570–600 for the beginning of decay, by associating the two groups of finds into a single phase of development to make a stronger date.
Dating summary (construction of plaza): range 500–525, midpoint 512.5, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (reconstruction of shops): range 530–55, midpoint 542.5, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (decay of plaza): range 570–600, midpoint 585, class Cs9 (contextual ceramics), Cs3 (associative, phase of development), publication 3/3.
15ORI Scythopolis: Two round plazas are drawn on a street plan in Agady et al., “Byzantine shops in the Street of the Monuments at Bet Shean (Scythopolis)”, in What Athens has to do with Jerusalem: Essays on Classical, Jewish, and Early Christian Art and Archaeology in Honor of Gideon Foerster (Interdisciplinary Studies in Ancient Culture and Religion 1) (Louvain 2002) 242, fig. 1 and most clearly on G. Avni, “’From Polis to Madina’ revisited—urban change in Byzantine and Early Islamic Palestine”, JRAS 21 (2011) (301–29) 304, fig. 1, which is what I will use here. They are set on the colonnaded streets running north-east of the city centre, the ‘Valley Street’ and that running NNW. Nothing is published apart from this plan, from which one can observe that the first plaza seems to measure ca. 68 m across from its rear portico walls, with porticoes of between ca. 7 and 8 m in width, and that the second plaza measures ca. 54 m wide from portico back wall to back wall, with porticoes of ca. 4 m in width. These measurements are highly unreliable, as they are hand-measured by me from what is a basic small-scale plan, intended only to show the basic organisation of the urban centre. In the first case they are shown in thick black lines, as if the dimensions had been confirmed by survey, whereas in the second case the lines were dashed, as if these walls were still highly conjectural and not confirmed by survey observations.
Dating summary: undated.
S3 Comprehensive Rebuilding of Fora / Agorai
06ITS Terracina: The forum of the city was restored (‘cleaned up’) by a dux Georgios, according to a Latin inscription written on a column now incorporated into the porch of the cathedral, which stands on the forum: ‘Mundificatus / est forus iste / tenp(ore) d(o)m(in)i Georgii / consul et dux.’ This text is dated to the late 6th or 7th c. (likely to be after complete pacification of Italy in AD 554, but before 661 when all dukes of Naples are known). As the inscription is inscribed on a column, the restoration of the portico is likely, but it might well have extended to the whole plaza. It occurs underneath a Greek acclamation, written by the same inscriber, wishing ‘Many years to the victorious and orthodox emperors’: M.C. Coppola, “Il foro emiliano di Terracina: rilievo, analisi tecnica, vicende storiche del monumento”, MEFRA 96.1 (1984) (325–77) 364–70; A. Guillou, “Inscriptions du duché de Rome”, MEFRM 83 (1973) 149–58 (who wishes to link the inscription to the presence of Constans II in 663, which I think is not necessary).
Dating summary: range 554–661, midpoint 607.5, class x (inscription not in situ), publication 3/3.
12CPL Constantinople (Augusteion): The plaza was restored by Justinian following a fire in AD 532, which destroyed porticoes as well as surrounding buildings, probably re-instating a basic shape that existed from AD 459 or earlier (see appendix K2b). Theophanes A.M. 6024 (AD 531/32 date in edn. of Mango and Scott (1997), although clearly January 532) describes the burning of the Great Church, St. Eirene, the hospice of Sampson, the Augusteion, the portico of the Basilica, and the Chalke; Malalas 18.72 describes the burning of the praetorium and Chalke, as far as the Scholae, the Great Church and the public colonnade [either a colonnaded street or forum portico]. See further references in appendix C3. We can reasonably assume that the square, which framed the Senate House restored by Justinian, was finished in time for the construction of the emperor’s column there, the set piece of Procopius’ description: Aed. 1.10.5–9. Theophanes gives AM 6036 (AD 543/44 date in edn. of Mango and Scott (1997)), 7th indiction, the 17th year of Justinian, for the erection of the column, a regnal year also given by Zonaras. It is given as the 16th year and 7th indiction by Cedrenos (edn. of Bekker (1838), vol. 1, p. 656). The safest TAQ for the works is 558. Procopius’ text (in which the plaza seems complete) likely dates to before the collapse of Hagia Sophia’s dome in 558, as acknowledgement of this does not feature in it, despite a description given of the monument in 1.1.66–78, which discusses problems of the stability of the dome, which would be odd given the expectations of the reader after 558. However, the speed of building at Hagia Sophia, of only 5 years from destruction to inauguration, suggests that 543/44 is a reasonable TAQ for the completion of the works and should be used, as specific local information is better than an argument based on rebuilding speed.
For the main summaries on this plaza see C. Mango, The Brazen House: a Study of the Vestibule of the Imperial Palace of Constantinople (Copenhagen 1959) passim; R. Janin, Constantinople byzantine: développement urbain et répertoire topographique (Paris, 2nd edn. 1964) 59–62 [now superseded]; R. Guilland, Études de topographie de Constantinople byzantine, vol. 2 (Berlin 1969) 40–54; W. Müller-Wiener, Bildlexikon zur Topographie Istanbuls (Tübingen 1977) 248–49; F.A. Bauer, Stadt, Platz und Denkmal in der Spätantike. Untersuchungen zur Ausstattung des öffentlichen Raums in den spätantiken Städten Rom, Konstantinopel und Ephesos (Mainz 1996) 148–67; A. Berger, Untersuchungen zu den Patria Konstantinupoleos (Ποικίλα Βυζαντινά 8) (Bonn 1988) 235–42. These provide more detail than I wish to explore here, with my views representing the current consensus, with the main references. My account is supposed to provide a basic discussion, with primary evidence, not review and comment on the theories of all previous scholars, which involve much speculation. See also K. Dark and J. Kostenec, “The Patriarchal Palace at Constantinople in the seventh century: locating the Thomaites and the Makron”, JÖB 64 (2014) 33–40 and K. Dark and J. Kostenec, “A new archaeological study of Hagia Sophia, Istanbul”, in Proceedings of the 22nd International Congress of Byzantine Studies, Sofia, 22–27 August 2011 vol. 3.1 (Sofia 2011) 213–37 for discoveries at the Patriarchate, which modify the topographical conclusions drawn by the summaries above.
Of plans, we have one showing undated walls, drawn casually, without stratigraphic recording, by Mamboury for C. Mango, The Brazen House: a Study of the Vestibule of the Imperial Palace of Constantinople (Copenhagen 1959) fig. 38, with detail in fig. 37. There is also W. Kleiss, “Beobachtungen in der Hagia Sophia in Istanbul”, IstMitt 15 (1965) (168–85) 150 fig. 1 (reproduced in Bauer (1996) 221 fig. 69), of which I saw only Bauer), both of which seem to have been drawn into J. Bardill, “The Palace of Lausus and nearby monuments in Constantinople: a topographical study”, AJA 101.1 (1997) (67–95) 70, fig. 2, with a few other walls of uncertain origin. Finally, there is a map currently being compiled by N. Westbrook (pers. comm. 2014) from maps of Mamboury in the German Institute, which he has very kindly let me see. It is worth remembering that the form of the plaza of the 4th c. may well have been different to that of the 6th c.: see esp. Mango (1959) 42–47; Berger (1988) 235–42, Bauer (1996) 153 and R. Stichel, “Sechs kolossale Säulen nahe der Hagia Sophia und die Curia Justinians am Augusteion in Konstantinopel”, Architectura 30 (2000) 1–25, with references to other literature. Bauer explains well a theory that the plaza was open to the west until Justinian closed it.
Of textual evidence, late 4th to early 5th c. authors (Socrates Hist. eccl. 6.18 and Sozom. Hist. eccl. 8.20) describe a controversial statue of Empress Eudoxia (excavated in the area of the square: appendix F9) as being located either one street width from Hagia Sophia or set before the Senate House. This suggests that there was no enclosed space in this area, whereas it was enclosed with 4 porticoes at the time of Procopius’ description: Aed. 1.10.5–9. This theory, on the early open shape of the plaza is, however, very speculative, depending entirely on texts. The only other concrete clue the texts give on the layout of the plaza is that the Column of Justinian / Augusteion was set at the west end [thus south-west corner] of Hagia Sophia (Petri Gyllii, De topographia Constantinopoleos, et de illius antiquitatibus libri quatuor (Lyon 1562) 2.17 pp. 102–103 and 2.18 p. 111 = J. Ball transl., Petrus Gyllius (Pierre Gilles), The Antiquities of Constantinople (London 1729) 2.17, p. 127 and 2.18 p. 137), on the left as one went into the south-west vestibule (two later medieval sources: Mango (1959) 47). The south-west corner must refer to the south-west corner of the church proper, rather than its atrium, to make any sense.
Of antiquarian evidence, Petrus Gyllius (2.18) also saw 7 Corinthian columns near to the pedestal of the column of Justinian, for which he gives measurements. These columns were also seen by C. Buondelmonti, Liber insularum Archipelagi (no place 1420) 65 (G.R. Ludwig von Sinner ed., Christoph. Bondelmontii, Florentini, Librum insularum Archipelagi (Leipzig and Berlin 1824) 122), who recorded them as 6 columns, next to the column of Justinian. Gyllius says that each of the columns was 30 foot and 6 digits high [30 m by 0.299 m + 6 by 0.0187 m = 9.0822 m], with a total colonnade height for column, capital and pedestal of about 46.5 foot [46.5 by 0.299 m = 13.9035 m], standing about 20 foot and 10 digits from each other [20 m by 0.299 m + 10 m by 0.0187 m = 6.167 m]. This massive colonnade, which is shown as 6 columns on some early drawings of the city, is thought by Stichel (2000) to represent the façade of the Justinianic Senate House (see below). If the colonnade did have only 6 columns this would represent a length of at least 37 m [6.167 by 6 = 37.002 m], although one might imagine a projecting porch with 4 front and two side columns [6.167 by 4 = 24.668 m] (as Procop. Aed. 1.10.6–9 describes for the Senate). In the latter arrangement, the columns would represent a projecting building façade and so need not represent the standard portico height of the rest of the plaza.
Of dimensions, the following suggestions can be made, for the 4 sides of the plaza:
(i) On the west side of the plaza a possible perimeter wall was recorded by Mamboury in the area marked G on his plan. This wall was also shown on the maps of Mamboury and Kleiss. It was found in 1934 and was 0.9 m thick and 12 m long, surviving to a height of 2 m above the original street level, and was adjoined at a right angle with another thicker piece of wall, extending to the west, which was aligned with the long axis of the ‘Basilica’ cistern: E. Mamboury, “Un nouvel element pour la topographie de l’antique Byzance”, AA (1934) (50–61) 54 (area F). This wall seems likely to represent the west perimeter of the plaza in Justinianic times; such a perimeter could not have been set any further west without blocking the major road beyond it.
(ii) On the north side of the plaza, another wall is drawn on the Bardill plan, that is shown turning a corner at its western end to align with a west ‘perimeter’ wall of Mamboury: J. Bardill, “The Palace of Lausus and nearby monuments in Constantinople: a topographical study”, AJA 101.1 (1997) (67–95) 70, fig. 2. This represents (p. 83 n. 67) a colonnade (of ‘square columns [i.e. piers] supported by simple marble bases’) at intervals of 3.4 or 3.5 m apart, found in 1962, 4.75 m inside the southern wall of the enclosure of Hagia Sophia, which ended 1.8 m from the western wall of the same enclosure, without any description of it turning a corner to the south here as on the Bardill plan, although the excavators suggest the corner might have existed at the end of the recorded portico. The masonry was similar to the shops on the Mese uncovered by the German Archaeological Institute and the Archaeological Museum of Istanbul: F. Dirimtekin, “Augustaion ve Milion”, IstArkMüzYill 8 (1969) 32–33 (English translation), 80. It is also important to note that the column bases of Dirimtekin’s portico were set at a height of 2.75 m below the ground level and 3.75 m below the ground level south of Hagia Sophia. The base of Eudocia / Eudoxia was found on the paving to the east of the plaza at the Palace of Justice at a depth of 3.5 m in 1847: Dirimtekin (1969) 31, though E. Mamboury, “Les fouilles Byzantines à Istanbul et dans sa banlieue immédiate aux XIXe et XXe siècles”, Byzantion 11 (1936) (229–83) 230 describes the paving here as being 3 m down. The height of the paving outside the Chalke Gate to the south-east of the plaza (the same building plot as where the Eudocia base was found) was recently recorded as 29.8 MASL in excavation compartment B according to G. Çiğdem, “La porte monumentale trouvée dans les fouilles près de l’ancienne prison de Sultanahmet”, Anatolia Antiqua 16 (2008) (259–90) 277 fig. 37, and is of a similar level (perhaps ca. 30.02 MASL) in compartment C according to p. 280 fig. 43, if one allows for 21 cm for each of the three steps, as in the compartment C drawing. Unfortunately, I do not have the absolute height measurements from the southern side of Hagia Sophia, to generate absolute levels here, but it seems that Dirimtekin’s portico was at the height of late antique paving in this area.
However, there are problems with this solution. This colonnade is unlikely to represent the north perimeter of the 6th c. plaza because it would make the Augusteion very narrow in its north-south measurement. It would also push the honorific column of Justinian too far away from Hagia Sophia to be considered ‘on the left’ as one went into the south-west vestibule of the church, as the column would have required at least 15 m clearance from this portico back wall, given the pyramid of steps that it had. Finally, the east side of the square, marked by the 6 column porch of the Senate House, seems, from antiquarian drawings (see below) to be sited between the Mausoleum of Sultan Selim II and the baptistery, so as to seem to form a backdrop to it. This would not be a reasonable location if the Dirimtekin colonnade was accepted as the perimeter wall of the complex. Mamboury located a number of structures in this area of likely Byzantine date (N. Westbrook pers. comm.), but they need not be part of the original Justinianic arrangement of the plaza.
An alternative, more northerly perimeter for the Justinianic square comes from the hypothesised edge of an audience hall of the Patriarchal Palace, recently calculated from observations of its preserved remains (surviving on the north and east side of the hall), revealed during plaster removal, which set its south-west boundary 3.5 m south of the baptistery of Hagia Sophia, aligning with the south wall of the baptistery: Dark and Kostenec (2011) 220–22, esp. 235 fig. 3. If this is correct (and the audience hall really was the southern boundary of the Patriarchal Palace), then the site of the Column of Justinian (foundation) would be located at the spot where E. Mamboury saw a mass of masonry standing to about 1 m above the ancient ground level, immediately to the right when leaving the south gate of the enclosure of Hagia Sophia (observed during work on drains in 1935). Mamboury himself interpreted this foundation as that for Justinian’s column monument: E. Mamboury, “Brickstamp Notes” (no place, no date) (not seen) referred to in Bardill (1997) 83, n. 67.
(iii) On the south side of the plaza, a new discovery allows us to set the position of the boundary wall. Recent excavations under the Ottoman Palace of Justice have been published in a summary report, which tells of a north-south boundary wall being discovered for the Great Palace, in which is set a portal 6.2 m wide, flanked by piers and other signs of monumental investment. This faces out via three steps onto marble paving, at a spot that is aligned with the ‘Milion’ sondage where a major junction of the Mese with the Regia towards the Chalke and the street running towards Hagia Sophia is indicated by a change in sewer direction: see appendix F3. An identification of this structure with the Chalke Gate is very reasonable, and it seems likely that the projecting piers are part of a tetrapylon built by Justinian, and that, like other tetrapyla built at crossroads, it fully filled the street junction and was integrated within the surrounding portico. If we project forward from it, we know the position of the Regia colonnaded street and its portico, the north portico rear wall of which must have formed the south portico rear wall of the Augusteion. This wall is likely to have projected out from the north-west corner of the Chalke Gate. If we allow the Regia the same width as the excavated ‘Chalke Gate’ piers, then it looks like we need to allow ca. 17.5 m for the street (porticoes of 5 m reflecting the piers’ width, separated by 7.5 m), and perhaps a further 1 m for the perimeter wall thickness.
(iv) On the east side of the plaza, things are more complicated. The north-south compound wall obviously sets a maximum distance east, within which the east perimeter wall must have existed, However, it seems unlikely that the east side of the Augusteion actually ran here, as a north-south street is recorded as existing here in the Middle Ages between the Chalke and the Holy Well on the south-east side of Hagia Sophia, alongside a raised passage running from the Chalke into Hagia Sophia: Mango (1959) 88–89. The street, at least, as a fundamental part of the urban framework of the area, is unlikely to have been carved out of the Augusteion and the Senate in the Middle Ages, but rather to have already existed in Late Antiquity. Thus, we should imagine the east perimeter of the plaza to have been located somewhere further to the west, beyond this north-south street. We can allow it the same width as given to the Regia, so ca. 17.5 m, with a further 1 m added for the perimeter wall of the plaza, and thus it is possible to set the maximum south-east perimeter wall position. The Senate House was located on the east side of this plaza, as described by Procopius and also Zos. 5.24, who notes it was ‘before the palace’, though the Senate was probably not equivalent to the palace hall of the Magnaura (Mango (1959) 56–57 vs. J. Kostenec, “Palace of Magnaura”, in Encyclopaedia of the Hellenic World, Constantinople (2008) http://www.ehw.gr/l.aspx?id=12436 (last accessed March 2016). Stichel (2000) points out the presence on antiquarian drawings (on 16th copies of an Album of David Ungnad of Sonnegg, of 1542–1600, shown on pp. 8–9, figs. 6–7) of 4 colossal columns aligned roughly north-south, just east of a water tower. This tower is known from old photographs to have stood near the south-east corner of the baptistery, but on the Ungnad drawings it is west of the tomb of Selim II. These columns are probably the same as the 6 columns described as being located between Hagia Sophia and the Hippodrome in a manuscript variant of 1466 of the description of the city by Cristoforo Buondelmonte of ca. 1420 and by several 16th c. writers: Stichel (2000) 2–3 with n. 14. These might represent the façade of the Justinianic Senate, with its 6 columns of white marble, described by Procop. Aed. 1.10.7–8 (described as being the biggest in the whole world), with 4 projecting forward, arranged as a prostyle porch, with what sounds like an arcuated pediment. Their position, set anywhere between the centre of the baptistery and the west wall of the Mausoleum of Sultan Selim II, would put the Senate façade (facing west) between ca. 75 and 90 m west of the north-south compound wall excavated with the Chalke.
Overall, taking all the recorded / hypothesised walls together, we can plot these points together to give a range of rough measures, for a plaza which seems to be slightly trapezoidal, as it tries to reconcile the different alignment of the palace compound wall found at the Chalke with that of Hagia Sophia, which is not set exactly perpendicular to it. (i) The north-south measurement of the plaza is clear: it begins with a portico back wall set 3.5 m south of the baptistery and ends with a portico back wall aligned with the Regia axis drawn out west from the Chalke (appendix F3) up to the point where the sewer changes direction in the ‘Milion’ sondage, at a major junction (appendix F3). This north-south measurement for the plaza gives us is ca. 690 m, if we give 1 m for the wall thickness to the south, incidentally giving us the length of the west portico. (ii) The east-west measurement of the Augusteion could be defined by the wall excavated in 1934 on the west side of the plaza and by the palace compound wall that joins the Chalke’s north-east pier. This east-west measurement is 151 m, from wall face to wall face, measuring off a modern accurate city plan provided by T. Öner pers. comm. But, we should not take the palace compound wall as representing the east portico wall of the Augusteion, as a north-south street separates it from the palace. This street, running north through the Chalke up to the south-east corner of Hagia Sophia, ought to be the same width as the Chalke’s north side. Its width can be reconstructed as 17.5 m, as the piers of the arch suggest as tetrapylon design, of which we know the east side dimensions (of 17.5 m, see appendix F7a). We should allow a further 1 m for the width of the Augusteion east wall. This would reduce the Augusteion east-west width to 132.5 m, from here to the wall of 1934, giving us the same measurement for the south portico length. However, if we take the columns observed by Stichel as the façade of the Senate, and imagine that this was set within an east portico, then the east-west plaza measurement shrinks to between 76 m and 61 m [151 m–75 m or 151 m–90 m] considerably.
The only major uncertainty is the nature of the Senate porch. Did it project into the plaza, leaving long paved areas between it and the north and south porticoes, as in an imperial forum at Rome? Or was the porch just marked by a pediment set within a wider portico, as on the palace façade at Arles, or in the depiction of the Palace of Theoderic from Ravenna (see appendix E1). We cannot know. We must allow at least 72.5 m [90 m–17.5 m] for the length of the Senate from the front of its porch stylobate to the exterior face of its back wall (taken as being 1 m thick), which served as the major back wall of the west portico of the north-south street between the Chalke and the Holy Well at the south-east corner of Hagia Sophia. We may we imagine this major back wall is also the back wall of the east portico of the Augusteion. In this hypothesis, the Senate must project into the plaza at least 71.5 m from the interior face of the east portico. This would have created a building about twice as long as the Diocletianic Curia in Rome (see appendix L1) but very close to the Aula Palatina in Trier. The Senate would have been flanked on each side by two narrow areas of paving leading up to the east portico, from which the building projected. But we can very easily imagine a second hypothesis, in which the façade porch projected into the plaza from its portico only by one or two columns. In this hypothesis, the Augusteion’s east portico stylobate would have been encountered ca. 6.167 m or 12.334 m east of the facade of the Senate (see Gyllius’ Senate porch column measurements above). If the portico back wall was 5 m beyond this, then under hypothesis 2 we can calculate an east-west plaza measurement of at least 79.5 m [61 m + 6.167 m + 12.334 m] and at most 94.5 m [76 m + 6.167 m + 12.334 m]. I take the first of these figures as being more likely, and use it here. This would give us lengths for the north and south porticoes, although the plaza is slightly irregular in its angles. of 82 m for the north and 78 m for the south.
From all this, it seems that we are getting close to a square plaza, in which the foundation of the honorific column of Justinian, at the spot identified by Mamboury, was central. This is the option I support here, based on the popularity of square plazas elsewhere in the period. If we accept Mamboury’s observation of the column foundation as being correct, we could envisage the following internal measurements for the plaza, taking account of its trapezoidal boundaries. I do this by hand-measuring off the city plan provided by T. Öner, onto which I have projected the alignments the street past the Chalke, of the Regia, the baptistery of Hagia Sophia, and the 1934 wall. These allow me to create east, south, north and west boundary markers for the square (without allowing for the Senate’s facade). This would suggest: (i) a west portico length of ca. 69 m, as above; (ii) a south portico length of ca. 78 m, as above (narrowed by the Senate porch to 61 m); a north portico length of ca. 82 m and an east portico length of ca. 71.5 m (new measurement), in all cases assuming boundary walls were 1 m thick. I take the columnar monument of Justinian, which seems likely to have reused a Constantinian stepped base that once supported a column of Helena, as being the same size as that of Constantine in his forum, so 12.4 m square. To cut a long story short, it seems that the north perimeter wall championed by Bardill is unlikely to be correct for the Justinianic period, and that the true north perimeter in Late Antiquity equated rather with the south side of the Patriarchate. Taken with the projected walls of the Regia on the south-west, the boundary wall on the north-west, and the known position of Justinian’s column and the Senate, we can imagine a plaza with a paved area of some 65 m square, without its porticoes, which may have been set so as to even out the divergent alignments of the site. All of this, is of course, highly speculative and rougher than I would like to admit. Re-excavation and a laser plan of the area is obviously needed to correct all measurements given here.
Of portico architecture, from its later 6th c. state, we have indications from texts, as well as the archaeological discoveries of Dirimketin, described above. The plaza appears in its fully restored condition in a description of Procopius of the Column of Justinian (Aed. 1.2.1–12), along with the new Senate building (Aed. 1.10.5–9, when the agora is described as a περίστυλος = surrounded by a colonnade). We cannot be sure how much of the porticoes were new, although the work was most likely of the same scale as Justinianic works surrounding the plaza, with the new Senate House, new Chalke Gate and new Great Church of Hagia Sophia. Note the suggestion of Bauer et al. (discussed in appendix K2b) that the ‘Great Gate of Meletes / Meletius’, which was the site of the shrine of Justin I and Justinian (if Anth. Pal. 1.97–98 can be believed) in AD 520, survived into the Middle Ages (Cer. 1.2 and 1.8), which suggests that some parts of the pre-Justinianic square were retained, supporting my suggestion that the porticoes were not necessarily replaced.
Of paving, scholars have often noted how in Patria 2.17 the ‘court of Hagia Sophia’ was cleaned up by Justinian, who ἐµαρµάρωσεν (‘covered it in marble’). It is important to note that the text does not say paved it with marble. The same word also occurs in the description of Justinian’s work on the Augusteion contained in Suda iota, 446 Ἰουστινιανός. From this language, Justinian might have added marble revetment rather than paving stones. Admittedly, stone paving was uncovered in 1847 when the Ministry of Justice was built: see E. Mamboury, “Les fouilles Byzantines à Istanbul et dans sa banlieue immédiate aux XIXe et XXe siècles”, Byzantion 11 (1936) (229–83) 230; the paving was found to be of dark stone slabs [i.e. likely sandstone rather than Proconnesian marble?] measuring ca. 0.3 m by 0.15 m (given as 1 ft by 0.5 ft), very carefully laid, so presumably in a geometric arrangement: Constantius I (Patriarch of Constantinople) and T.M. Aristoklès, Kωνσταντίου Α´τοῦ ἀπὸ Σιναίου ἀοιδίµου Πατριάρχου Κωνσταντινουπόλεως τοῦ Βυζαντίου βιογραφία καὶ συγγραφαὶ καὶ ἐλάσσονες ἐκκλησιαστικαὶ καὶ φιλολογικαὶ καί τινες ἐπιστολαὶ τοῦ αὐτοῦ ἐξεδόθησαν µετὰ παραρτήµατος (Istanbul 1866) 381–84, not seen by me but cited in C. Mango, The Brazen House: a Study of the Vestibule of the Imperial Palace of Constantinople (Copenhagen 1959) 60. For sandstone paving elsewhere in the centre of the city see appendix C3 on the Lower Mese phase 1 (of possibly Severan date). However, as discussed in appendices C3 and F3, the excavations of the Ottoman Palace of Justice seem to have uncovered the site of the Chalke on the very western edge of the site, where a major boundary wall for the palace was pierced by a wide opening (6.2 m), flanked by two piers, which aligns perpendicularly with the southern end of the Mese, where there is a major corner in a sewer that angles towards the Chalke. This makes it unlikely / impossible that the dark stone slabs are part of the Augusteion. Here, recent excavations have rather brought to light three steps leading off the boundary wall just south of the arch and then marble slabs, likely relating to a street going south from the Chalke. These slabs also provide an indication of the likely covering of the Regia leading past the square, and also of the Augusteion: A. Denker, G. Yağcı and A.B. Akay. “The Great Palace excavation”, in Istanbul: 8000 Years Brought to Daylight, edd. A. Karamani Pekin and S. Kangal (Istanbul 2007) (124–41) 135–36. We seem, however, to have no direct evidence for the paving of the plaza, to date.
Of statuary, apart from the honorific column described in appendix F9, we hear that Justinian’s redevelopment of the Augusteion involved extensive statue removal (including late antique imperial statues, such as the columnar statue of Theodosius that this column replaced). ‘427 statues’ were relocated, of which 80 were of Christians: Parastaseis 11 (Patria 2.96). This occurred after catastrophic damage to the statues of the Augusteion in a fire of 475 according to Suda mu 120 Malchus, which notes that Malchus’ history especially lamented this. Perhaps Parastaseis (copied into Patria) is confused here and the clear-out happened before Justinian, under Zeno or Anastasius. Given that many imperial statues of earlier centuries survived into the Middle Ages on the adjacent colonnaded street of the ‘Regia’ and the ‘Street by the Milion’ (see appendix H7), this suggests that the Augusteion after Justinian would have felt noticeably different to the surrounding spaces, being dominated by this imperial image. Just as under Anastasius, when the Mese had been cleared of its bronze statues, it seems the Augusteion and perhaps the ‘Basilica’ were cleared of damaged art works after the fire of 532, or perhaps before. On the identities of the statues removed, see appendix K2b and appendix F9 for the columns, though that of Eudoxia seems to have survived the Justinianic cull, as its inscribed base was excavated in the area of the plaza in the 19th c. (see above).
Dating summary (for restoration of the plaza by Justinian): range 532–44, midpoint 538, class x (historical texts), publication 3/3.
12CPL Constantinople (Forum of Constantine): We should anticipate some repair after a fire in the forum of Constantine in AD 582/83: Theophanes AM 6075 (AD 582/83 date in edn. Mango and Scott (1997)). I take this repair work to have occurred within 5 years of 583, given the completion of the new Hagia Sophia by 537.
Dating summary (repair after fire): range 582–88, midpoint 585, class x (historical), publication 3/3.
12CPL Constantinople (Forum of Theodosius): The Forum of Theodosius shows signs of a second phase that affected the whole plaza. The extant eastern arch was reduced to a slimmer two-piered portal, with the destruction of its outer pillars. The eastern portico rear wall of the square west of the arch overlies one of these external piers, so revealing that much of the present state of the plaza is of the second phase: R. Naumann, “Neue Beobachtungen am Theodosiusbogen und Forum Tauri in Istanbul”, IstMitt 26 (1976) (117–41) 130ff. with fig. 131. Based on Naumann’s plan, the north wall of the square (Naumann (1976) 122–23 (figs. 4–5 detail plan and site plan)) seems to be of a different (primary?) phase: it does not align with the northern terminus of the rear wall of the east portico, although the junction between the two areas is not excavated. For the identification of these remains as the Theodosian Forum, which is reasonable based on the find-spots of fragments of what seems to be the column Theodosius, and based on the surrounding terraces, see: A. Berger, “Tauros e Sigma. Due piazze di Constantinopoli”, in Bisanzio e l’Occidente: arte, archeologia, storia. Studi in onore di Fernanda de’ Maffei (Rome 1996) (17–31) 18–19.
Of phasing, the second period of construction seems to involve a rebuild of the interior colonnade and rear wall of the eastern portico in mortared rubble alternating with 5 layers of bricks, which involved demolishing the external piers of the ‘Arch of Theodosius’. It is also likely that works involved a rebuilding of the colonnade with stylobate of the northern portico in mortared rubble, as described in appendix K1a. The dimensions of the porticoes do not seem to have changed, the north portico being ca. 13.7 m wide (from front of external stylobate to back wall) and 76 m long; the east portico being ca. 11.25 m wide (from front of external stylobate to back wall) and 85.25 m long. The north portico external stylobate has no column bases, but rather an alternation of rectangular and square emplacements, suggesting perhaps an alternation of piers and columns, an arrangement which was not entirely regular at its eastern end. The excavation plan of Naumann (1976) 122 fig. 4 shows that an exit in the rear wall of the new east portico was paved in marble slabs, although they were rougher rectangular slabs of slightly different dimensions, of which one was possibly reused. The slabs measured ca. 1.2–1.6 m long by ca. 0.7–0.9 m wide, not like the straight-cut and mathematically-arranged marble slab paving in the forum north portico and plaza, which thus look to be part of phase 1. These measurements were derived from hand-measuring off the plan. These rougher slabs are likely to have been lain across the whole rebuilt portico, though Naumann fig. 4 labels a plain ‘Estrich’ inside the portico, which may be the underlay of the slabs, which have been robbed elsewhere.
Of dating, Mamboury believed the second phase work in the eastern portico to be of the third quarter of the 6th c. (mentioned by Naumann (1976) 125, based on an archive plan of Mamboury’s 1943 excavation). This is probably because of the construction style of mortared rubble, which, in the part of the walls observed by Naumann in 1969/73, were of mortared rubble interspersed with 5 layers of bricks. Bauer would like to date this work to after the AD 599/600 earthquake [not in Guidoboni Catalogo], which damaged the forum: F.A. Bauer, Stadt, Platz und Denkmal in der Spätantike. Untersuchungen zur Ausstattung des öffentlichen Raums in den spätantiken Städten Rom, Konstantinopel und Ephesos (Mainz 1996) 190–91 citing Anonymous Chronicle on Constantine, esp. 97–106. This explanation should be disregarded, as we cannot know physically what restoration work, if any, followed this event. The most credible TPQ for the works comes from the incorporation of one of the distinctive tree-trunk columns from the area of the arch (perhaps from an attached portico, as it is not big enough) into the Basilica Cistern under Justinian, after the riots of AD 532. Given the speed of the works on Hagia Sophia (inaugurated in 537) we might allow 5 years for the completion of the rebuilding of this very prominent public monument on the city’s main monumental avenue. See also my appendix C3 for more detail on the secondary phase of the Upper Mese between the Forum of Theodosius and the Forum of Constantine.
Dating summary: range 532–37, midpoint 534.5, class Cs4 (reused material presence), Cs5 (catch-all masonry), x (historical), publication 3/3.
12CPL Constantinople (‘Basilica’ courtyard): This enormous courtyard was rebuilt for the second time, after the fire of AD 532. The central court of the ‘Basilica Cistern’ was paved and porticoes of the ‘same Basilica’ were built by urban prefect Longinus (PLRE 3.795–96), recorded as prefect between 536 or 537 and ca. 542 (see PLRE 3.1479 Fasti for other urban prefects, of 536 and 543): Malalas 18.91. A few years earlier, Justinian had built a cistern inside the very large court at the ‘Imperial Portico’ (where lawyers and judges prepared their cases): Procop. Aed. 1.11.12. According to Procopius this cistern went under the ‘portico which faced towards the south’, a statement which implies its demolition and rebuilding. See also Chron. Pasch. Olympiad 327 AD 528 (dating from edn. of Whitby and Whitby (1989)); Theophanes A.M. 6020 (AD 527/28 date in edn. of Mango and Scott (1997); I follow Chron. Pasch. for the date here, given its earlier composition date). Given the speed of the works on Hagia Sophia (inaugurated in 537) we might allow 5 years for the completion of the rebuilding of this very prominent public monument. Earlier, it had been restored by consul Illus in 478, who gave his name to it: see appendix K2b. As described in the main text, the plaza was a centre of legal activity and the home of the city’s library.
For recent summaries see F.A. Bauer, Stadt, Platz und Denkmal in der Spätantike. Untersuchungen zur Ausstattung des öffentlichen. Raums in den Spätantiken Städten Rom, Konstantinopel und Ephesos (Mainz 1996) 219 nn. 11–12; W. Müller-Wiener, “Yerebatan Sarayi”, Bildlexikon zur Topographie Istanbuls (Tübingen 1977) 283–85; C. Mango, The Brazen House: a Study of the Vestibule of the Imperial Palace of Constantinople (Copenhagen 1959) 42–47, 48–51; R. Guilland, Études de topographie de Constantinople byzantine (Berlin and Amsterdam 1969) vol. 2, 4–13.
Of dimensions, the size of the courtyard is often conjectured by using the preserved extent of this large ‘Basilica Cistern’ / Yerebatan Sarayı as the basis for the courtyard, and then co-opting various aligned walls drawn on the map of Mamboury, on his plan in C. Mango, The Brazen House: a Study of the Vestibule of the Imperial Palace of Constantinople (Copenhagen 1959) fig. 38, and especially a long wall running parallel to the north-east side of the cistern, which looks like the boundary wall of the complex, shown on Bauer (1996) 221 fig. 69 (reproducing W. Kleiss, “Beobachtungen in der Hagia Sophia in Istanbul”, IstMitt 15 (1965) (168–85) 150 fig. 1, of which I saw only Bauer) and drawn into J. Bardill, “The Palace of Lausus and nearby monuments in Constantinople: a topographical study”, AJA 101.1 (1997) (67–95) 70, fig. 2. The Basilica Cistern’s great size (138 m by 64.6 m in Bauer (1996) 220) provides some guide to the dimensions of the plaza before the reconstruction by Justinian, which it ante-dated by a few years. A cross-check with Kleiss, measuring off his plan, suggests that 64.6 m is the external width from the outer edge of the boundary wall, whereas 138 m is the internal length from the inside of the boundary wall, the external length of which is actually ca. 145 m.
Of dimensions of the Justinianic structure, we know from Procopius that the cistern filled the court inside the complex and part of the ‘portico which faced towards the south’. It presumably had to damage this portico, because the court was not big enough on its own to hold the cistern. The walls shown on the map above give us the northern boundary and also the eastern boundary of the complex (with less certainty, as the wall is known for only a few metres). These allow us to envisage a structure that had a much wider northern than eastern portico. We do not know if the structure was symmetrical, but assuming that it was, we can come up with an estimate of the internal foundations. A problem is that the cistern itself is not precisely aligned with the surviving perimeter of the basilica courtyard, making it hard to give consistent measurements. Given the presence of linear foundations within the northern basilica, aligned with the perimeter wall, it seems unlikely that the stylobate of the portico was carried directly on the perimeter wall of the cistern. The identity of the ‘portico which faced towards the south’ is problematic. For Bauer (1996) 220 this means the north portico, but Bardill (1997) 79 n. 51 believes it means the south-west portico (i.e. the south portico in this discussion). The latter is more likely, as the surviving walls suggest a very wide north portico beyond the end of the cistern, though admittedly I am here forcing texts and archaeology to agree.
If we measure on the plan of Kleiss, from the inside of the northern perimeter wall of the portico to the edge of the northern edge of the basilica cistern, then we get a portico width from the back wall to front of stylobate of ca. 22–22.5 m (north-east corner to north-west corner, which we will take as being 22.25 m average), hand-measuring roughly off the plan, whereas the same measurement for the eastern portico is ca. 12 m, for the small section where it can be measured. If the corrected cistern measurements (derived from Kleiss’ map) are followed, and we allow for a plaza that had a west portico as wide as the east, and a south portico as wide as the north, then the plaza would have measured ca. 109.1 m north-south (64.6 m + 44.5 m) by 169 m east-west (145 + 24 m), measuring from back portico wall to back portico wall. The plaza was certainly no bigger than this, as it would have encroached on the Mese and its shops if it had been (N.B. see my reinterpretation of the ‘Milion’ remains as the terminal end of a street portico in appendices C3 and F3). If the southern porticoes had been built above the cistern, following Bardill’s reading of Procopius, then the north-south width measurement could be up to 22.25 m less, so 86.85 m wide, which would give us the minimum possible internal dimensions for the plaza, of 86.85 m by 169 m east-west.
Of porticoes, Procopius (Aed. 1.11.12) describes the Basilica as being a great court surrounded on 4 sides by a περίστυλος, which the translation of Dewing takes as ‘columns’, but perhaps describes the sensation of being in a deep portico with more than one colonnade. Procopius says immediately afterwards (Aed. 1.11.13) that the same court is surrounded by stoas on 4 sides, which makes this interpretation likely. As noted above, at least the northern portico must have had a double or triple colonnade, or have been more like a traditional civil basilica, with a wider nave or a perimeter colonnade. Aligned walls (?stylobates) within the northern portico on the map of Kleiss, cited above, suggest one colonnade close to the northern wall, then two parallel colonnades set close together within the centre of the portico. It is difficult to tell if these really are stylobate foundations, or if they might represent drains or other features.
Of paving, we have some idea from observations by Mamboury (1936) 247, who saw large Proconnesian marble slabs of different lengths and widths, from 0.7 m to 1.05 m by 1.5 m to 2 m.
Of decoration, in the mid-5th c., the ‘Basilica’ courtyard is described as ‘gilded and adorned with marbles’: Cod. Iust. 8.11.21 (AD 440). Such gilding was likely reinstated by Justinian, as later sources describe it as having a gilded roof, which probably means gilded bronze tiles: Parastaseis 37; Patria 2.41. The adjacent Milion is called by Parastaseis 38 the ‘golden (ὠρέῳ) Milion’, which might refer to a golden milestone, but might also refer to a roof of gilded tiles.
Of statues, references were assembled by S. Bassett, The Urban Image of Late Antique Constantinople (Cambridge 2004) 153–56, all from medieval sources. These relate to mythological / animal statues, such as Parastaseis 37 / Suda beta 157 Βασιλική (Heracles); Patria 2.40 / Suda Βασιλική beta 157 (Solomon with hand holding chin / Justin I on bended knee, perhaps really a philosopher on account of the pose); Parastaseis 37 / Suda Βασιλική beta 157 (Elephant set up by ‘Severus’). It is possible that this group was established after the rebuild by Justinian, given that it includes no imperial statues of the 4th–5th c. as one might expect. It is safest to consider it only post-Justinianic. It could be later. Justinian’s workers notably cleared out a large number of statues from the Augusteion and may have done the same here: see entry above on the Augusteion. The replacement of the paving of the courtyard, covering the cistern, would of course have presented an opportunity to set out a new statue display, or at least a reconfiguration of the old one. The Patria claims a statue of Heraclius in the ‘Basilica’ courtyard: Patria 2.41 although Parastaseis 37 has it as the ‘measure’ of Heraclius, so some sort of grain measure. Parastaseis also lists a number of statues of late 7th to 8th c. emperors, which may be misunderstandings, or reinterpretations of ancient statuary, unless a residual tradition of statue dedication survived here in the very centre of Constantinople.
Dating summary (construction of cistern and hypothetical rebuilding of south portico): 528, class x (historical text), publication 3/3.
Dating summary (restoration of porticoes and paving after fire): range 532–37, midpoint 534.5, class x (historical text), publication 3/3.
12CPL Constantinople (Strategion): We should anticipate some repairs after a fire in the Strategion, recorded by Marcellinus Comes in AD 510, which was followed by the immediate restoration of the arm of [bronze] statue holding a horn of plenty, set above an arch. I take this repair work to have occurred within 5 years of 510, given the completion of the new Hagia Sophia by 537 after the fire of 532, but it is possible it was all done in 510, the same time as the statue.
Dating summary (repair after fire): range 510–15, midpoint 512.5, class x (historical), publication 3/3.
13ASI Ephesus (Tetragonal Agora): This enormous square courtyard complex is made up of 4 porticoes, each 11.2 m wide, surrounding a courtyard of 111 m by 111 m, off which opened around one hundred cellular one-room units, which were most likely built as shops. It was comprehensively rebuilt in the later 4th c. (ca. 379–403: see appendix K2b), and then again in part in the later 6th c. For a summary with full bibliography see G. Aristodimou, “Ephesus (antiquity), commercial agora (Tetragonos)”, in Encyclopaedia of the Hellenic World, Asia Minor, URL: <http://www.ehw.gr/l.aspx?id=8214> (last accessed March 2016). The complex has only been published in preliminary form, especially for Late Antiquity, so what follows here can only be regarded as a very provisional summary, partly based on my site observations. The latest statement on the agora is P. Scherrer, “Die Agora: Vorläufiger Baugeschichte”, in Forschungen in Ephesos XIII.2 Die Tetragonos Agora in Ephesos, edd. P. Scherrer and E. Trinkl (Vienna 2006) (49–54) 48–51, replacing the first report of W. Wilberg and J. Keil, Ephesos III.1 Die Agora (Vienna 1923), but even the new report of Scherrer is by no means a thorough exploration of the phases and characteristics of the late buildings. For a plan see Scherrer (2006) 339 fig. 5.
Of late modifications, of 6th c. date, there are the following elements: (i) a late marble floor of ‘mid to later 6th c. date’; (ii) the construction of a new 2 to 3 storey north portico (which involved blocking of the shops behind); (iii) a partial rebuilding of the east portico shops, in association with works on the Hall of Nero above; (iv) the laying of a new southern entrance ramp; (v) the creation of a central building known as the ‘horologion’. See below, appendices S5a [for (i)], S4 [for (ii)-(iii)], S6 [for (iv)] and S10a [for (v)].
Of phasing and dating, P. Scherrer, “Die Agora: Vorläufiger Baugeschichte”, in Forschungen in Ephesos XIII.2: Die Tetragonos Agora, edd. P. Scherrer and E. Trinkl (Vienna 2006) 51–52 (49–54), sees this building work [apart from the east portico rebuilding, which I identify myself] as part of a coherent phase of occupation of the agora. He presents no evidence for this apart from a common higher floor level for the north portico, which is apparently reflected in the west portico shops. However, it is possible that the later marble floor [i.e. paving of the open area of the plaza?] is also set at a common higher level, although this is not stated and no independent dating evidence is given for it. He would like to date the overall phase by association with the last major ancient phase of occupation revealed in one of the west portico shops: a glass workshop, which has produced coins of the time of Justinian, which supports my own dating of the north portico to the period 550–610, based on the use of spolia and its relationship to the adjacent late fortification: S. Karwiese, “Agora: ein Münzteppich”, in S. Karwiese et al. “Ephesos”, ÖJh 67 (1998) Grabungen 1997, 11–12. Other parts of the phase do have their own independent dating, which is not far off this range. The date of the ‘horologion’ building, which contains reused material at the centre of the agora, depends on late 5th c. coins from layers associated with / underlying its construction (which give it a rough contextual date of 487.5–512.5, under the rules of this study), see: Scherrer (2006) 52. The southern entrance ramp, built of reused material, sealed a coin of AD 544/45, see S. Karwiese, “Das Südtor der Tetragonos Agora in Ephesos”, ÖJh 66 (1997) 306, 315. Clearly, there are a lot of repairs, but how much is part of a unified rebuilding is open to debate. There are at least arguments from the common use of rare spolia pieces to suggest that a partial east portico shops rebuild belongs at the same time as the north portico construction, so qualifying the work undertaken as an ‘agora restoration’, as this requires two elements to be listed here, under the rules of this study.
I argue elsewhere (appendix C3) that the articulation of the ‘Byzantine City wall’ with the Arcadiane suggests that it was built at the same time as the creation of the artificial ‘Police Station’ Hill behind the northern portico, and that this was part of a unified urban development before the time of Phocas and certainly of Heraclius, under which emperors the gate in the fortification was inscribed with acclamations. These acclamations were part of a set that spread along the Marble Street to the south. As the acclamations were clearly started under Phocas, a date prior to the end of his reign in 610 provides a TAQ for this larger urban redevelopment, whilst a tetrastylon established in the middle of the Arcadiane provides a TPQ of ca. 550, on grounds of artistic parallels, given the tidier reuse associated with it, which is much messier in the work connected with the city wall. In support of this date, we can note the presence of reused blocks derived from the Prytaneum, material that was also found in the buttressed rear walls of the north portico / its adjoining street wall and also in contexts around the Baths of Scholasticia of 6th c. date, such as the Curetes Stoa of ca. 550. There are also different pieces of a single inscription, found in both the portico wall and around the Baths of Scholasticia. See appendix K2b for details, with the possibility that spolia use also indicates a rebuilding of the eastern portico of the agora (perhaps just its vaulted shops) at the same time. Further research is clearly needed on the agora to better fix these phasing observations, and to retrieve ceramic dating. The ramp and horologion at least might date to a little later than the 610 limit I have placed on the north portico, perhaps as late as 614 (the Persian invasion of Asia Minor), after which civic secular building is not known.
Dating summary: range 550–610, midpoint 580, class Cs3 (associative, phase of development), Cs7 (TPQ coin), Cs6 (TAQ inscription), Cs5 (catch-all masonry, spolia use), publication 3/3.
15ORI Antioch in Syria (implied): The rebuilding of the agorai of Antioch, after the complete destruction of the city in the Persian sack of AD 540, is implied by Procop. Aed. 2.10.19–22, who records the re-establishment of the street grid of the city. He describes paving slabs each big enough to fit in a wagon, laid down in the ‘level / cleared ground of the city’. I interpret Procopius’ description of the works as indicating, or at least including, the agorai of the city. This seems reasonable, as immediately after this quotation, Procopius says that stoas [i.e. colonnaded streets] and agorai were then constructed. One could note that the excavator Lassus did not find slabs of this size in the Justinianic phase of the main street, from his limited sondages: see appendix C3. However, stones big enough to fit ‘one to a wagon’ are known on other late antique agorai, as in the Forum of Constantine at Constantinople. The agorai of Antioch did survive into this period, as Evagr. Hist. eccl. 1.18 describes a number of buildings known to be located on the Hellenistic Agora as still existing at his time of writing (late 6th / early 7th c.). Justinian must have respected these structures in his rebuilding work. On the history of the Hellenistic Agora of Antioch see the analysis in G. Downey, A History of Antioch in Syria from Seleucus to the Arab Conquest (Princeton 1961) 624–31. It is difficult to establish a TAQ for the completion of Justinian’s rebuilding work, but I will here use 25 years after the sack of 540, under the rules of this study, which takes the range for the building up to 565, the death of Justinian.
Dating summary: range 540–65, midpoint 552.5, class x (historical), publication 3/3.
15ORI Scythopolis (‘Byzantine agora’): A trapezoidal agora had been built sometime in the 5th c., overlying some temples (appendix K1a). It was remodelled in the second phase, of early 6th c. date, with a shortening to a more rectangular shape, by the creation of a new portico: Y. Tsafrir and G. Foerster, “Urbanism at Scythopolis-Bet Shean in the fourth to seventh centuries”, DOP 51 (1997) (85–146) 122–23 with a plan on fig. D. See the report of G. Mazor and R. Bar-Nathan, “The Bet She’an excavation project— 1992–1994: the IAA expedition”, Hadashot Arkheologiyot—Excavations and Surveys in Israel 105 (1996) (7–36) 17–20. Other features (listed below) were added at the same time, amounting to a major refurbishment of the square.
Of dimensions, the plaza now had a maximum length of ca. 85 m, from back wall to back wall, through the introduction of a new portico on the south-east side, without shops, ca. 46 m long and ca. 5 m wide, from back wall to front of stylobate. I took all but the first of the measurements by hand off a small-scale version of the plan, the first being from the published text above, which also gives the width of the north-west side as being ‘more than 50 m’, which I took by hand as ca. 54 m. The measurement of 85 m given by the text concurs with my hand measurements, suggesting that the latter are fair. Thus, I am able to discount the dimensions given by Mazor and Bar-Nathan (1996) 17 for the agora, of 83.5–84.5 m by 116–122 m. These seem to be perimeter wall measurements from the first phase. However, the same authors give a second phase length for the agora of ca. 88 m on p. 18, which is likely to be a correct measurement from back wall to back wall, more accurate than my ca. 85 m above. Given the confusion, I stick to my own measurements of ca. 54 m by ca. 85 m for the second phase.
Of porticoes, I was able to see on site that the new (south-east) portico had a stylobate of limestone ashlar blocks, but that no details of the colonnade elevations were surviving, although column shafts were strewn around the site: L. Lavan site observations 1998. Three types of portico paving were recorded, of which one was phase 1, whilst two and three both seem to belong to phase 2. The first floor was of mosaics decorated with depictions of animals and plants (p. 122), which I was able to see on site (in the south-east corner of the reduced square) as being the first type of paving, slightly lower than the second paving (L. Lavan site observations, April 1998). I could see that a mosaic inscription was set into this mosaic carpet, which must be that to the governor Marcianus (Tsafrir and Foerster (1997) 122). This was a large-scale unified polychrome mosaic carpet with a border of plants featuring this inscription, and also what seems to be a tiger. This mosaic floor was replaced in the south-east corner of the 6th c. ‘reduced plaza’ by a floor of bi-colour (‘black’ and ‘white’) square tiles. These were laid in a geometric pattern, which one might consider opus sectile, in a grid parallel to the stylobate in the south-west portico and as a grid running diagonally in relation to the stylobate of the south-east portico: L. Lavan site observation, April 1998. Here, a report of 1996 clarifies, with photos, that an opus sectile pavement in limestone tiles was laid, for which see Mazor and Bar-Nathan (1996) 18–19 fig. 18. Elsewhere, within the porticoes, there was a third type of paving: an oil shale / bituminous limestone pavement, for which see Mazor and Bar-Nathan (1996) 18–19 and Tsafrir and Foerster (1997) 122. On the north-east side of the agora, over the former basilica, was found a marble slab, embedded into this pavement, that was inscribed with a Greek text recording that the governor Rometalkes [perhaps better Romytalkes, with thanks to Gian Luca Gregori, La Sapienza, Roma], a scholasticus, undertook ‘The whole work of paving the portico and of the building’: G. Foerster and Y. Tsafrir, “City center (north): excavations of the Hebrew University Expedition”, Excavations and Surveys in Israel 11 (1992) (The Beth Shean Excavation Project 1989–91) (3–32) 7.
Of entrances, a new path paved with basalt slabs led up into the agora: a limestone block that was set into the pavement carried the following inscription: ‘In the time of Flavius Romytalkes, the archon and the scholasticos (the restoration of this word is unsure), all this work was done’: transl. G. Foerster and Y. Tsafrir, “Center of ancient Bet Shean—north”, in: “The Bet Shean project”, Excavations and Surveys in Israel 6 (1987–88) (25–35) 31.
Of shops, this phase apparently saw many changes to the shops in the west portico, in their layout and their floors, “some of which follow a style characteristic of the 6th century”: Nazor and Bar-Nathan (1996) 18.
Of internal features, it is probable that the stepped monument / large honorific column base at the centre of the plaza (which perhaps supported a statue or cross) also dates from the same time as this portico, or a little later, as it is set at what is almost the centre of the reduced plaza, but not the centre of the original larger plaza: see appendix F9.
Of phasing, there is some difficulty in drawing together all the elements of renewal into a single phase, as the site has not yet been fully published. The east portico wall does provide evidence of a plaza-wide plan, as it is connected to adjacent floors. The monument at the centre of the agora has clearly been placed there to suit the new shape of the agora after the east portico was built, which means it should be contemporary or later than these developments. The shale floor and ramp clearly belong together, as they have inscriptions of the same governor. However, I have not been able to confirm that there was a mosaic beneath the oil- shale paving slabs, as there was beneath the black and white tiles. Yet, it seems likely that the two non-mosaic floors were part of the same renewal, with the black and white tiles perhaps completed under a different governor in oil-shale slabs. Probably, the two floor types in the porticoes belonged together, in the second phase, as the excavators have suggested. In the absence of evidence to the contrary, I will accept this.
Of post-antique phasing, an Umayyad potter’s complex overlies the centre of the south-east portico, disturbing all except its last few metres at each end, as well as extending into the plaza. This produced jugs and asymmetrical flasks in yellowish clay, typical of the Umayyad occupation levels at Scythopolis: Tsafrir and Foerster (1997) 138 with plan fig. D. However, during the occupation of this period (which begins in 635), some of the shops were rebuilt on the south-west side of the plaza, which is shown on the phasing of plan D (as a row of similar sized one-room cellular units), as well as being referred to in the text of Tsafrir and Foerster (1997) 138 (N.B. they mistake the side of the agora, giving the location of the shops as the north side, rather than the south-west side). I was able to confirm this, on site: basalt shop walls set at a higher level did overly the earlier shops, with one unit having a (higher) floor that seemed to be made from basalt rubble: L. Lavan site observations, April 1998.
Of dating, Tsafrir and Foerster (1997) 122–23 say that “the chronological framework of the two phases is not clear, but it seems likely that the second phase of building falls in the time of Anastasius”. They consider that the agora was “remodelled at the beginning of the 6th c.”, giving no evidence in justification, though referring to Y. Tsafrir, “Scythopolis capital of Palestine secunda”, Qadmoniot 107–108 (1994) (117–31) 130–31 (not seen). However, it is possible to extract chronological information from the published reports. A broad, if not very useful TPQ for both phases of the plaza can be given based on other information I have: this square was built partly over the early Roman basilica, and is given a TPQ by ‘5th c.’ coins and pottery in robbing trenches of this structure: Tsafrir and Foerster (1997) 122–23. A strong TAQ for both phases of the plaza can be calculated from the inscription of the governor Romytalkes, for which see above. The governor does not claim to be a consul, and so the inscription should date before AD 529–35. This is because the governor of Palaestina Secunda was a praeses, the lowest grade of civil governor, at least until AD 529 (PLRE 3.177 Bassus 3) but not later than AD 535, when a governor is named as a consul in an inscription from the city: G. Mazor, “City center of ancient Beth Shean—south”, in “The Bet Shean project”, Excavations and Surveys in Israel 6 (1987–88) (10–23) 17. This is Flavius Nyssios Sergios the Scythopolite, dated on the specific year 598 = AD 535, given by the inscription, [which can be revised to 534/35, though this does not affect our dating]. This elevation in rank for the governor might perhaps be connected to the raising of the adjacent province of Palaestina Prima to proconsular status in 536: Iust. Nov. 103. Given that the plaza was severely disrupted by the Umayyad occupation, it is unlikely that any of the monumental investment of this phase took place in the square in the 7th and early 8th c., especially in the south-east corner, where the limestone bi-colour tiles existed. Thus, we can set a tentative TPQ for the late antique phase 2 of AD 635. This is reasonable, as both the governatorial inscriptions and the central monument suggest that the work involved the self-representation of late antique governors and emperors in a traditional format, which would not suit a time under the Islamic occupation.
Of dating, for the Umayyad phase, the Potter’s Complex established here was apparently destroyed in the earthquake of AD 749 [Guidoboni Catalogo no. 220]: R. Bar-Nathan and G. Mazor, “City centre (south) and Tel Iztabba area: excavations of the Antiquities Authority Expedition”, Excavations and Surveys in Israel 11 (1992) (The Beth She’an Excavation Project 1989–91) (33–52) 37. For evidence of damage and dating of this site-wide earthquake, see appendix A11. A recent report has revealed that building 1 of the Potter’s Complex, built across the south-east portico of the square, contained a hoard of 122 silver dirhams of between AD 707 and 747, whilst a lintel from the building was inscribed in Greek with the Arabic formula ‘Allah is the only one’: R. Bar-Nathan and A. Najjar, “Urban pottery workshops in the Umayyad period: Beth She’an and Jerash”, in R. Bar-Nathan and W. Atrash, Bet She’an II: Baysān : The Theater Pottery Workshop (IAA Reports 48) (Jerusalem 2011) (191–202) 196. The end date for the coin hoard at least fits within the chronology of the site-wide destruction. We are not told by the report what the context of the hoard was, i.e. a buried treasure or a box / pouch of coins, although such contextual information would be significant, as a hoard in open use (such as a cashbox) that ends in 747 could be used to support the extension of the ‘earthquake destruction of 749’ to this area.
Dating summary (for refurbishment of agora): range 500–35, midpoint 517.5, class 0 (no evidence), Cs8 (contextual coins), Cs6 (TAQ inscription in situ), publication 1.5/3 (as TPQ evidence not published).
Dating summary (for Umayyad occupation phase): range 635–749, midpoint 692, class 0 (no evidence), Cs3 (associative, ceramics, as I have not seen contexts), Cs2 (catch-all other, earthquake), publication 2/3 (as pottery and contexts not published in the report I had access to).
See also (appendix S4): 07AFR Carthage (Maritime Agora): Rebuilding work within the period 533–58 may have been extensive, with new porticoes and a new entrance arch, if the Admiralty Island Round Plaza is the correct identification for the maritime agora of Procopius. 10MAC Philippi: Rebuild of the south portico is accompanied by at least partial rebuilding of the east and west porticoes, as part of the same phase of development. The rebuilding of two or more elements at the same time qualifies this as forum / agora comprehensive rebuilding, which is dated to between 536 and 616. This is a more complex phasing than the ‘troisième etat’ offered by M. Sève and P. Weber, Guide du forum de Philippes (Sites et monuments 18) (Athens 2012) 22–26, with 30–31. One does not necessarily need to believe that the additions to the ‘basilica’ / ‘Hall of 5 columns’ (see appendix X1b) were constructed at the same moment. The presence of a series of diagonally-aligned reinforcement piers in the south-east corner of the plaza, at the junction of the south and east porticoes, reveals that phases subsequent to the rebuild of the south and east porticoes could be made, although in the same type of masonry. Rather we need structural arguments, like the presence of L-shaped piers or design similarities, such as the similar design of porticoes in their piers, to justify grouping similar but isolated masonry into the same phase of construction.
S4 Porticoes, New or Rebuilt
07AFR Carthage (Maritime Agora): New stoas were built by Justinian on either side of the ‘maritime agora’, recorded by Procop. Aed. 6.5.10–11. As more than one stoa was built, I treat this as a comprehensive restoration. This work should date between the reconquest of 533 and the collapse of the dome of Hagia Sophia in 558. The latter event is not referred to in Procopius’ text, despite a description given to the monument in Aed. 1.1.66–78, which discusses problems of its stability. We cannot be sure if the ‘maritime agora’ he refers to relates to the circular plaza on the Admiralty Island of the circular harbour, although this is at least possible, and the plaza certainly has 6th c. repairs: see appendix K1a and esp. V4b.
Dating summary: range 533–58, midpoint 545.5, publication 3/3.
10MAC Philippi (south portico) (2: construction and repair): A new arcaded portico was built along the south side of the forum, of a length of ca. 100 m, using Ionic impost capitals, as part of the third phase of the forum: M. Sève and P. Weber, Guide du forum de Philippes (Sites et monments 18) (Athens 2012) 22–26, with 30–31. See also M. Sève, “L’oeuvre de l’École française d’Athènes à Philippes pendant la décennie 1987–1998”, Αρχαιολογικό Έργο στη Μακεδονία και στη Θράκη 10b (1996) (705–15) 708–709; M. Sève, ‘Le forum de Philippes’: http://users.otnet.gr/~provost/Forum/Forum.html (last accessed 01/03/2003, no longer exists), fig. 3 for plan.
Of dimensions, the portico measures ca. 100 m long and 14.16 m wide (divided equally into two naves by an internal colonnade), according to Sève and Weber (2012) 51. These are measurements that I could confirm against plan 6 on p. 22, which revealed that the length of ca. 100 m is taken from the side of the portico facing the forum plaza, whilst the width of 14.16 m is that from back wall to plaza-side front of stylobate. However, if one measures the length of the portico along its back wall, one can add a further 13 m in length, necessary for the join of this structure with the other porticoes of the plaza.
Of architectural form and features, the south portico was entirely rebuilt: (i) the steps down onto the plaza were altered (only the lowermost step was original, with three above being re-laid); (ii) the stylobates were re-laid; (iii) the Corinthian colonnade of the façade was substituted for one with monolithic smooth columns and Ionic bases (reused from the previous portico), but with new-cut Ionic impost capitals, designed to support an arcade; (iv) the interior colonnade was replaced by a series of rectangular masonry piers, set wide apart: Sève and Weber (2012) 51, who provide no measurement for colonnade heights. I was able to add a few details from my site observation of April 2017. The portico forum colonnade’s west corner pier [which is L-shaped and served both the south and west porticoes] was composed of rubble, reused stone and brick (irregularly placed within the masonry), apparently uncoursed. The reused material was of a small size. The same portico colonnade’s east corner pier [which is rectangular and serves only the south portico] in contrast uses a form of opus mixtum, with bands of two bricks and rubble, with a very small amount of fragmentary reused material. This might be a fragment of an earlier phase, unrelated to the late rebuild. Onto this rectangular pier a new part was added to make it into an L-shaped pier. It was composed of rubble and broken brick fragments, without brick coursing. Other parts of the portico here were of the same rough masonry type, notably the east end of the final intermediate colonnade stylobate of the south portico (with some large squared blocks set at the east edge), the piers of the said internal colonnade (with rather larger blocks and a bit of reused material), and the final piers of the adjacent east portico. The south portico internal stylobate was created initially by building mortared rubble walls between the original pier emplacements: L. Lavan site observations April 2017.
Of internal dimensions, the outer colonnade had an interaxial measurement of between 2.65 and 2.85 m. The internal colonnade divided the naves into bays of 3 m to 3.15 m, with the piers being on average 0.93 m long and 1.65 m to 1.75 m wide. Both colonnades varied in their measurements internally, both between the piers and between the porticoes. The last bays at each end of the internal colonnade were wider, to allow for north-south access here, out of and through the portico. At the ends of the external colonnade, facing onto the plaza, the two junctions with the east and west porticoes were reinforced with an L-shaped pier, composed of mixed masonry, for which the masonry facing survives in one case: Sève and Weber (2012) 23, with p. 51 for the lowermost step of the stairs and p. 53 for the detail of the L-shaped pier’s masonry facing. Using an interaxial median of 2.75 m and photos from my site visit in June 2017, I am able to estimate very roughly that the columns used were ca. 3 m high and the bases ca. 0.33 m. I was not able to estimate the heights of the impost capitals or any arcades.
Of phasing, phase 3 of the forum is based on a common masonry type. This is made up of more irregular blocks with quite disparate material (lots of reused material, bricks, pebbles), bound together with solid mortar that is not homogeneous, containing sand and little stones in variable quantities. There are sometimes gaps in the walls, which have been filled with mortar. These ‘mediocre’ walls are covered with a thick layer of plaster: Sève and Weber (2012) 30–31. However, we are not always told exactly which walls belonged to this phase, and phasing is often just presented as fact, rather than being justified with wall descriptions. Admittedly, in the case of the portico stylobate, we are told that there is a mixture of disparate elements including an abundance of reused material, but such descriptions are not frequent, nor clearly related to the phase description: Sève and Weber (2012) 23. It is difficult not to think that ‘phase 3’ of the forum is an oversimplification, rolling up all the latest visible monumental repairs on the plaza into a single construction period. Sève believes that all the work in phase 3 belongs together, in a comprehensive programme of rebuilding, after an earthquake [not identified, though see candidates in Guidoboni Catalogo p. 688–704]. This is a seductive hypothesis for the basilica and curia area: here reinforcements between the colonnades can be seen, and post-marks support the curia façade on the forum paving, suggesting earthquake damage. Furthermore, the southern portico was built on a deep fill, which would only be necessary if it were to hide debris, whilst the shops to the south lost their complicated upper-storey elements. These changes suggest that a seismic event is credible. However, it is also possible that the expensive repairs that made up phase 3 might have taken longer to complete than Sève thought, extending some decades after the earthquake, into the 6th c.
Of dating, the presence of impost capitals in the south portico seems to be the only basis for Sève’s chronology for the phase 3 repairs to the forum, which he gives as ca. AD 500 in old publications, and as late as the 5th to the early 6th c.: Sève (2012) 22. However, this dating seems to be too early on stylistic grounds. The capital from the portico shown by Sève (2012) 22–23 fig. 7 corresponds well with those in the site of Aliki (on Thasos, opposite Philippi) in the basilique Sud (V. Vemi, Les Chapiteaux ioniques à imposte de Grèce à l’époque paléochrétienne (BCH Supplement 17) (Athens 1989) 168 no. 224 with pl. 65, no. 224 = J.-P. Sodini with J. Servais, Aliki I : les deux sanctuaires. Les carrières de marbre (Études Thasiennes 9) (Paris 1980) 135 with fig. 118 and pl. 47b). Another capital from the Philippi portico (now on top of a column in the portico, measuring 0.37 m high) is shown by Vemi (1989) 159 no. 196 with pl. 57, no. 196, which is also known from Aliki (basilique Sud), although with its scrolls absent: Sodini with Servais (1989) 134 with fig. 117 and pl. 47a.
Thus, the dating of the basilique Sud of Aliki becomes critical as a basis for dating the capitals of Philippi’s south portico. The presence of “tribunes” (i.e. galleries) in this same basilique Sud are a feature in the region from the time of Justinian, based on dated parallels (pp. 136–37), and suggest a date in the later 6th to early 7th c. However, a sherd was also retrieved from a destruction layer (layer 3, from the trench between the two basilicas), associated with the construction of the basilique Sud, which directly covers the fill of its foundation cut. This sherd was a local ceramic, imitating Late Roman C Hayes 10 A2, which is given as dating to the end of the 6th and the beginning of the 7th c. [LRC Hayes 10a is given as 570–612.5 in Atlante]. The end date of 619 seems to be based on the last dated coin in a coin hoard from Thasos, marking the start of a downturn attested across the island: O. Picard, “Trésors et circulation monétaire à Thasos du IVe au VIIe siècle après J.-C.”, Thasiaca (Paris 1979) (BCH Supplement 5) 451, 454 (not seen). Similar coin hoards of the later 6th and early 7th c. across Greece and Macedonia attest to similar ruptures elsewhere in this period: A. Avramea, Le Péloponnèse du IVe au VIIIe siècle: changements et persistances (Byzantina Sorbonensia 15) (Paris 1997) 72–80. Whilst this is not a very strong TPQ, the whole of the evidence, especially the sherd of Late Roman C, suggests a date for the ‘basilique Sud’ of Aliki in the late 6th to earlier 7th c., for which 619 is a reasonable local terminus for the widely-attested regional downturn. J.P. Sodini notes (pers. comm. 2016) that the sculpted form of the capital could represent a type that existed already in the 6th c. Aliki need not be the primary model. But the existence of not one but two parallels to Aliki is compelling.
Overall, we have to admit that we do not have very strong dating for the south portico of the forum of Philippi. However, the stylistic parallels with Aliki for the capitals provide the best dating so far. Before being surprised by a later 6th c. date, we should remember that there were two major ecclesiastical constructions on adjacent parts of the forum at this date. Both the second Octagon in the episcopal complex and Basilica B were obviously inspired by Hagia Sophia (completed in 537) and SS Sergius and Bacchus (completed in 536), both in Constantinople (Basilica B especially so, leading it to being given a date of ca. 540). The south portico of the forum ‘links’, if indirectly, to the Octagon by a ‘forum phase 3’ doorway and corridor, established in the library in the south-east corner of the portico. The second Octagon also has a colonnaded atrium, accessed via its entrance, which uses impost blocks, although they do not carry decorated capitals. An arcaded street portico with similar impost capitals, on the Via Egnatia, even post-dates this new colonnaded atrium (see appendix C4). Perhaps the rebuilt south portico of the forum actually dates from a similar time to these surrounding ecclesiastical developments, being designed to serve them, by embellishing the through route between the churches, along the south side of the forum. Admittedly, the slightness of the ‘link’ to the Octagon, seemingly an afterthought rather than a great piece of monumental connection, suggests that the portico was rebuilt before the second Octagon (sometime between 536–616, see appendix E1). However, the spatial association of the new portico with the two churches and, above all, the capital parallels to Aliki, suggest that a date in the period 536–616 is plausible.
Of subsequent phasing, there is the installation of reinforcement piers / buttresses in the east corner of the south portico. Here a diagonally-angled reinforcement pier, also of uncoursed stone with brick fragments, was inserted behind the L-shaped corner pier. Indeed a series of similarly-aligned reinforcement was built at the junction of the south and east porticoes, set around the pier to the south-east in the internal colonnade and the perimeter wall beyond it, being of the same construction, reinforcing the junction of the porticoes here right across the whole of the join: L. Lavan site observations, April 2017. There is no means of distinguishing the chronology of this later repair from the late construction of the south portico.
In terms of a TAQ for the whole sequence, the last coins from the forum and from the adjacent Octagon church are of the time of Phocas (reigned 602–10, although the coins are given as 605–10), when the neighbouring Octagon is believed to have been destroyed by an earthquake: Sève and Weber (2012) 26 [nearest known option is that recorded in Constantinople in 611: Guidoboni Catalogo no. 203]. However, for the sake of consistency I have used 616 as TAQ. This is the date of the definitive loss of control to the Slavs in the north Balkans, after which secular civic public building is not known in the region.
Dating summary (construction and repair of late south portico): range 536–616, midpoint 576, class Cs1 (architectural style), Cs3 (associative, phase of development), z (regional development), publication 3/3.
10MAC Philippi (East portico): The inner colonnade seems to have been rebuilt, with a mortared rubble ‘phase 3 masonry’ stylobate supporting rectangular piers of mortared rubble, one of which includes a broken architectural fragment that looks like a small piece of fluted column. Otherwise, there is no obvious reused material or brick, as far as I could see, until one gets towards the north end, where the colonnade is replaced by a solid wall that has some bricks and also possible reuse (a variety of blocks, some well-squared) alongside rubble: L. Lavan site observations, April 2017. Furthermore, a brick and rubble extension to the east corner pier of the external colonnade of the south portico was made, to turn it from a rectangle into an L-shaped pier, serving the east portico. This development reflected the L-shaped pier at the west end of the south portico in its late construction phase. This suggests that the rebuild in this area is part of the same phase of works as the south portico rebuild, which the late piers resemble in all their characteristics. As such, I give it the same date, based on phase of development, on account of the style of its capitals and on the surrounding architectural context.
Dating summary (construction and repair of late south portico): range 536–616, midpoint 576, class Cs1 (architectural style), Cs3 (associative, phase of development), z (regional development), publication 3/3.
10MAC Philippi (west portico): The internal colonnade of the west portico (south half) has piers of the same masonry as the south portico rebuild, of rubble, reused stone and brick (irregularly placed within the masonry), apparently uncoursed, set on a stylobate of the same type, set above the previous solid ashlar column emplacements. The piers are blocked by an ?unmortared rubble wall in two intercolumniations as far as I can see, with no tiles visible. Transverse blocking walls in the portico include rubble and some reused material: L. Lavan site observations, April 2017. The L-shaped pier at the south end of the portico is obviously shared with the south portico in its late construction phase. This suggests that the rebuild in this area is part of the same phase of works as the south portico rebuild, which the late piers resemble in all their characteristics. As such, I give it the same date, based on phase of development, on account of the style of its capitals and on the surrounding architectural context.
Dating summary (construction and repair of late south portico): range 536–616, midpoint 576, class Cs1 (architectural style), Cs3 (associative, phase of development), z (regional development), publication 3/3.
13ASI Ephesus (Tetragonal Agora): Excavators believe that the north portico of the agora, which was built at a higher level than the rest of the plaza, dates to the 6th c.: P. Scherrer, “Ephesos (Türkei) 1.1 Agora”, in “Jahresbericht 1999 des Österreichischen Archaeologischen Instituts”, ÖJh 69 (2000) 13–16; P. Scherrer et al., “Ephesos (Türkei) 2.2 Tetragonos agora”, in “Jahresbericht 1999 des Österreichischen Archaeologischen Instituts”, ÖJh 69 (2000) 374; Scherrer P. et al., “Ephesos (Türkei) 2.3 Tetragonos agora”, in “Jahresbericht 2000 des Österreichischen Archaeologischen Instituts”, ÖJh 70 (2001) 258 and especially the final report P. Scherrer, “Die Agora: Vorläufiger Baugeschichte”, in Forschungen in Ephesos XIII.2 Die Tetragonos Agora in Ephesos, edd. P. Scherrer and E. Trinkl (Vienna 206) (49–54) 50–52, along with P. Scherrer ed., Ephesus. The New Guide (no place 2000) 140–46. A site plan is available from Scherrer (2006) 339 fig. 5 with room numbers, whilst his fig. 4 on p. 338 reproduces the site plan and reconstructed section of west. Wilberg and J. Keil, Ephesos III.1. Die Agora (Vienna 1923), the original report, which still contains some useful observations and photos on the north portico on pp. 12–13.
Of dimensions, the portico occupies a similar size of space to the other three porticoes of the agora, so doubtlessly follows those of its predecessor. The length is ca. 135 m (measuring along the rear wall) and the width ca. 11.5 m (from back wall to front of stylobate), roughly hand-measuring off the plan of Wilberg. Using this measurement, combined with site photos taken during a site visit in April 2013, I was able to roughly estimate (from the last column with capital and base to the east on the internal colonnade) the following dimensions: base ca. 0.45 m, column ca. 3.55 m, capital ca. 0.75 m. This last colonnade does not have an impost capital surviving above, but adjacent columns do support them (of a likely height of ca. 35 cm, with some larger) making a combined colonnade height of ca. 5.1 m for the lower storey.
Of architectural form, the above reports, especially Scherrer (2000) and (2006), describe the structure as being a 2 or 3-storey double-naved arcaded portico. A second storey, at least, is indicated by the presence of stairs at the western end of the portico: Scherrer (2006) 51 (niche 3 in his plan 5); Wilberg and Keil (1923) 13 with fig. 17, showing the staircase well-preserved. The new stylobate is composed of reused column fragments and architrave pieces and it cuts the Augustan-Tiberian and also later 4th c. foundation (“Punktfundament”, presumably the foundation of the earlier stylobate): Scherrer (2006) 51. The new rear wall is a heavily buttressed construction. This wall is for a short distance, at the western end, composed of ashlar piers (likely shop portals of the original north portico shops) with inter-pier spaces blocked by coursed rubble and tile construction. These piers only last for 3 or 4 ‘shops’, however. For the rest of the wall’s length, the buttresses and the ‘blockings’ are part of a single heavy retaining wall, composed of the same material, of large reused architectural blocks of ashlar, which entirely replace any shops that once stood here: L. Lavan site observations, April 2003 and April 2013; Wilberg and Keil (1923) 12–13 with fig. 17; Scherrer (2006) 50–51 with fig. 69. A latrine was installed in one of the surviving shops at the west end of the portico: Scherrer (2006) 52, whereas three rooms at the east end of the portico were preserved, though their front wall onto the portico was reinforced with benches and piers: Scherrer (2006) 52 with fig. 74. Of these piers, that at the very end of the portico ended in an impost block, and so must have supported the vault of the portico: L. Lavan site observation, April 2013.
Of architectural features, we have few published details of the colonnade elevations. The internal colonnade is the best preserved, of which the lower storey has been re-erected by the Selçuk Museum: Scherrer (2006) fig. 12. It is composed of grey smooth columns and one fluted column. See my calculations above for their size. From the remains found on site, we can say that the columns were set at an interaxial spacing of 3 m and were mostly unfluted, set on early Roman bases, with impost blocks supporting Corinthian capitals that carried brick arcades: Scherrer (2006) 51. I noticed that the bases were Attic, although there were some quarry-ready bases in different parts of the colonnade, which might have been new-cut in Late Antiquity: L. Lavan site observation, April 2013. The capitals are thought to have been taken from the temenos colonnade of the Serapeion, being a little smaller (0.77 m rather than 0.79 m in height) than a similar capital used in rebuilding work on the middle colonnade of the Hall of Nero: Scherrer (2006) 52, fig. 58. They are also published in Wilberg and Keil (1923) 5–16, figs. 21–24.
Of phasing, it is important to note that the rear wall retained a new artificial hill, as did an alley way leading from the north portico towards the top of the Arcadiane, creating a higher area bounded on the north side by the ‘Byzantine city wall’, which formed the rear wall of the street portico of this stretch of the Arcadiane. This work seems to represent a major levelling operation, which seems from my observations to be part of a single programme of works connected to the creation of the ‘Byzantine city wall’ (before AD 610, see appendix C3). Scherrer (2006) 51–52 considers this building is part of a general phase of occupation of the agora (including the late building at its centre), for which see the discussion above. The new north portico also cuts through walls of the earlier late antique ‘Theodosian’ phase.
Of dating, the above phasing suggests a date of 550–610. In support of this date, we can note the presence of reused blocks derived from the Prytaneum, material that was also found in the buttressed rear walls of the north portico / its adjoining street wall and also in contexts around the Baths of Scholasticia of 6th c. date, such as the Curetes Stoa of ca. 550. There are also pieces of the same inscription, found in both the north portico wall and around the Baths of Scholasticia. See appendix K2b for details, with the possibility that patterns of spolia use also indicate a rebuilding of the east portico of the agora (perhaps just its vaulted shops) at the same time. Furthermore, a 6th c. date for the north portico can apparently be suggested, based on the impost capitals: Scherrer cites a pers. comm. of E. Russo on p. 51 n. 233, but no archaeological dating evidence can fix it securely. Scherrer also sees a common 6th c. floor level to the agora, in both the new portico and the shops. He believes this is dated by a 6th c. ‘Münzteppich’ [coin-spread] in one of the shops (West Stoa chamber J), which contains coins of Justinian as its latest issues. A late entrance ramp also sealed a coin of 544/45. For this information see appendix S3.
Dating summary: range 550–610, midpoint 580, class Cs3 (associative, phase of development), Cs3 (associative, coins), Cs6 (TAQ inscription), Cs5 (catch-all masonry, specific spolia), publication 3/3.
13ASI Aphrodisias (South Agora, west portico): The west portico of the south Agora was restored, outside the Hadrianic Baths. A long series of acclamatory inscriptions (ALA 83), cut onto 20 columns of the portico, claim that it was built by the clarissimus Albinus, a private benefactor, who is called ‘builder of the stoa’ by ‘the whole city’.
Of phasing, we are forced to admit that, whilst not part of the primary phase of the plaza itself, the portico shows no sign of reused material in its main columns (unlike the south portico, which was also rebuilt). The west portico also incorporates a ‘fronton’ from the adjacent Hadrianic Baths, suggesting that it perhaps belongs in the Hadrianic period: N. de Chaisemartin and A. Lemaire, “Le Portique de Tibère: recherches sur son architecture et sa fonction”, in Aphrodisias Papers 3, edd. C. Roueché and R.R. R. Smith (JRA Supplementary Series 20) (Ann Arbor, Michigan 1996) (149–172) 155–56 with fig. 7. C. Roueché in ALA 83 describes it as “Nineteen out of an original twenty white marble columns, all on identical bases. The columns, ca. 0.15 in diameter [which must be a typing error— L. Lavan], are smooth to a height of 1.73 and fluted above. They are made in two sections; the join between the two is not always at the same point, being sometimes below and sometimes above the line where the fluting begins, but the overall effect is uniform and elegant.” The portico does incorporate some reused material in the foundation of its stylobate: de Chaisemartin and Lemaire (1996) 154 fig. 6 with 150 fig. 1. However, this may not be a sign of dating to Late Antiquity, as reused material in foundations happens throughout Antiquity. Thus, it is not yet entirely clear what the late building of the portico involved: perhaps it was limited to a re-erection of the columns, some paintwork, or a roof repair.
Of statuary, the works done were substantial enough to see a base for an honorific statue erected by the city to Albinus, who is identified as a native of the city and an evergete. The fragmentary coloured marble columnar base bearing this inscription was found in the west portico, at the southern end: ALA 82 = LSA. See also the map of statue bases and finds in R.R.R. Smith, “Statue life in the Hadrianic Baths at Aphrodisias, AD 100–600: Local context and historical meaning”, in Statuen in der Spätantike, edd. F.A. Bauer and C. Witschel (Wiesbaden 2007) (203–235 with pls. 50–68, figs. 1–42) fig. 42 (B45) with commentary on p. 228 and 231.
Of the acclamatory inscription, see ALA 83 with C. Roueché, “Acclamations in the later Roman empire: new evidence from Aphrodisias”, JRS 74 (1984) 181–99. The acclamations were well-inscribed, with traces of red paint in the letters, and were distributed across each column in a sequence, following the order in which they would have been acclaimed. Roueché thought that a date in the 6th c. is preferable. She provides a lot of complex arguments, with parallels in the use of language (e.g. a verb and the epithet φιλόπατρις) to other honorific texts of the early or middle 6th c., from Aphrodisias and Beirut. She believes that “the odder elements of the script, such as the cursive delta and mu” fit best with the 6th c. She noted that the use of clarissimus for someone outside of the Senate would also seem to be appropriate after the mid-5th c. However, Roueché has recently noted that the overall dating of acclamatory inscriptions of this type seem to be from the time of Justinian (527–65) up to the early 7th c., thus modifying the range of the Albinus texts [which incidentally involves passing over the fact that the emperors are plural in the inscription]: C. Roueché, “From Aphrodisias to Stauropolis”, in Wolf Liebeschuetz Reflected, edd. R. Salway and J. Drinkwater (London 2007) (183–192) 186. This affects the wider dating for Albinus, suggesting that the statue also belongs in the same period. I now adopt a TAQ of 614 for all of these works, the date of the decisive Persian invasion of Asia Minor, after which no civic secular public building is known.
Dating summary (portico restoration): 527–614, midpoint 570.5, class Cs6 (absolute, inscriptions in situ), publication 3/3.
Dating summary (statue base Albinus): 527–614, midpoint 570.5, class Cs6 (absolute, inscriptions in situ), publication 3/3.
13ASI Aphrodisias (South Agora, south portico): For the most detailed descriptions, see N. de Chaisemartin and A. Lemaire, “Le Portique de Tibère: recherches sur son architecture et sa function”, in Aphrodisias Papers 3 (Journal of Roman Archaeology Supplementary Series 20), edd. C. Roueché and R. R. R. Smith (Ann Arbor, MI, 1996) (149–172) 157–58, who believe that the portico was made entirely out of spolia. This study contains earlier bibliography, and its authors report a sondage at the east end of the portico, revealing that it overlay an Early Imperial predecessor. A brief survey was also carried out by L. Crema, “Monumenti architecttonici Afrodisiensi”, in “Gli scavi della missione archeologica italiana ad Aphrodisiade nel 1937”, MonAnt 38 (1939–40) (233–312) 288–92, with fig. 60 (reconstruction drawing). There are also comments in P. Stinson, Aphrodisias 7: The Civil Basilica (Wiesbaden 2016) 83–85 with plan on p. 84 fig. 43.
Of dimensions, we have precise figures from Crema (1939–40) 292, who notes that the columns are set 6.95 m from the wall. However, these figures are problematic. If we take them to mean that the columns begin 6.95 m from the wall, then the back wall to front of stylobate measurement may be larger. The map of C. Roueché in ALA online is also of little use, as it contains no scale, which would have permitted measurements to have been estimated for vertical and horizontal dimensions: http://insaph.kcl.ac.uk/ala2004/images/iPlans/0001/02.gif (last accessed August 2016). Fortunately, I was very kindly given an up to date (2015) site plan with a scale by R.R.R. Smith, to whom I am grateful. From this, I could see that the portico measured ca. 7 m wide at its eastern end (from back wall to external front of stylobate) and was ca. 8 m wide at its eastern end (from back wall to external front of stylobate), hand-measuring off the plan. The portico’s length along its back wall is ca. 221 m long, but along the plaza side it is ca. 211 m (as the back wall runs into the west portico). A plan of the west end of the portico is also available in Stinson (2016) 83 fig. 43, on which the portico width, from the back wall to the front of the stylobate, was 7 m where it passes in front of the civil basilica, and ca. 7.3 m where it passes in front of a street, where late shops have been built. The figure I will use ‘officially’ for the portico width is ca. 7–8 m. From Smith’s plan, I could also see that an opening with stairs at the rear wall of the portico, towards its eastern end, was ca. 4.625 m wide (i.e. it was somewhere midway between 4.5 and 4.75 m, hand-measuring from the pixelated plan). This last critical measurement permitted some observations (below) to be made from photos I took in 2003.
Of phasing, there is evidence of late rebuilding at both ends of the portico, although the regularity of the spacing in the central section, alongside the fairly homogeneous elements found in some parts of the portico, clearly indicates that we are dealing with the rebuild of an Early Imperial portico, not an entirely new structure. At the west end, inscriptions have been found recording late rebuilding, which is also suggested by the irregular spacing of the columns here (see below). At the east end, a sondage has revealed a late ‘stylobate’ composed of a crude ?gravel, onto which a reused square block was set to support a column of the portico: de Chaisemartin and Lemaire (1996) 157 fig. 8. At the east end, the same authors also notice two sets of beam holes in the back wall of the portico, where it backs onto the rear of the theatre: p. 157. It is here that the two phases of the portico are clearest. I can confirm this from site observation based on photos I took in 2003, using the critical measurements derived from Smith’s plan. However, I saw series of holes (1 row built cavities, 2 rows large holes, 2 rows small holes).
Of these rows, there seem to be two or three systems. Firstly, the line of built rectangular cavities is the most obvious, and likely earliest. These occur ca. 10 m above the ground level, set at a height where ashlar masonry stops and coarse stone blocks of much smaller size begin. The row is formed by the regular setting of small ashlar blocks leaving gaps between each block. Into each block is set a hole, making up one of the rows of small holes. These two lines of holes seem likely to represent the roof structure of the Early Imperial portico, of which we do not have any other obvious traces in the wall. Secondly, there are two lines of widely-spaced large holes at ca. 6.5 and ca. 9.25 m above the assumed ground level. These two sets of holes correspond with each other, but are irregularly spaced and irregularly cut. The heights are not always the same either. They are set at intervals that seem likely to coincide with columns, with each column defining a bay by supporting a floor beam. However, the interval of the columns is not that of the extant portico as it has survived / has been re-erected (I have not been able to confirm if they are in situ). I took these two rows to represent the floor beams for a room set above the portico architrave and for its attic, within a system that reused the first system of holes for rafters. Thirdly, a first row of closely-set but irregularly-spaced small holes, ran at a height of one row of stone blocks above the large-hole’s ‘1st floor’ floor joists (so ca. 7.25 m above the assumed ground level). This row was very straight, but the holes in it were irregularly spaced. This may be a floor / ceiling system associated with the first hole system or could relate to a different late antique rebuild than that visible in the second hole system. Given the flimsy and irregular character of the last row of holes I think it is likely that we are in the presence of two late antique rebuilds. However, no certain evidence elsewhere in the portico confirms a second late antique phase, so I err on the side of caution, and stick with one late rebuild.
Of architectural features, the portico is not very homogeneous, although I noticed, at the eastern end, that it has similar column types, made of ‘long drums’ / sections, fluted over half their length, probably reused from the previous portico. De Chaisemartin and Lemaire (1996) 157 provide a description of the architectural elements of the colonnade. The bases are of many types (classic Attic, Attic with upper torus in speira, Hadrianic “bases à gorge”, late simplified bases, two bases on an “orthogonal” socle, “des colonnes 3 et 4 en partant de l’ange sud-ouest”): de Chaisemartin and Lemaire (1996) 157. The study of Crema (1939–40) 294–95, with 298 fig. 57 and 301 fig. 60 reveals that bases 3 and 4 are ‘octagonal’ [not ‘orthogonal’]. He also publishes photos of ‘unfinished’ pedestal bases in the portico, p. 295 fig. 56. I myself noticed two Ionic capitals, ornamented with wreaths, in place on columns near the east end of the portico: L. Lavan site observations 2003. Some late new-cut bases and columns (having a “schematic form”, rougher than the reused elements) are apparently present in the colonnade: Stinson (2016) 85.
The entablature seems to have featured an architrave rather than arcading. At the western end, an architrave block with a late inscription was recovered (see below). A second architrave is mentioned by Crema (1939–40) 291, as being similar to those of the northern portico, although a bit smaller, like some other elements of the structure. This block was likely taken from a ruinous portico on this site, originally harmonious with the northern portico. However, the architrave was not long enough for all of the interaxial distances between the columns. A third architrave, thought to be part of an original portico, is confirmed as reused by Stinson, studying the point where the portico fronts the civil basilica: P.T. Stinson, “The civil basilica: urban context, design, and significance”, in Aphrodisias Papers 4, edd. R.R.R. Smith and C.J. Ratté (JRA Supplementary Series 70) (Portsmouth, R.I. 2008) (79–106) 102 n. 60.
Of colonnade height dimensions, from my very rough calculations, off a photo, using the method outlined above, I would estimate the columns of the portico at 5 m in height: L. Lavan site observation 2003. Stinson notes that the new portico columns at the east end of the south portico would have reached the top of the central door of the basilica’s north façade: Stinson (2016) 85. This column height can be calculated as 6 m, if one hand-measures from his reconstructed section drawing of the façade on P. Stinson, “Local meanings of the civil basilica at Aphrodisias: image, text and monument”, in Basiliques et agoras de Grèce et d’Asie mineure, edd. L. Cavalier, R. Descat and J. des Courtils (Mémoires 27) (Bordeaux 2012) (107–26) 123 fig. 14, though the late portico floor height is not explicitly made clear on this drawing. This compares with my estimate of 5 m for columns at the east end of the south portico, although we do not know if bases and capitals are included in Stinson’s 6 m figure.
Of colonnade plan dimensions, the interaxial measurements between columns are most regular towards the west: de Chaisemartin and Lemaire (1996) 157. At the far west end of the portico, the large ‘floor beam’ holes in the first ‘early portico’ sequence correspond well in their positions with the columns of the extant portico, suggesting that the late reconstruction had been conservative: L. Lavan site observations 2003. However, the interaxial measurements oscillate wildly between 2.35 m (8th to 9th column from west) and 2.79 m (16th to 17th column), and the columns do not align well on the stylobate: de Chaisemartin and Lemaire (1996) 157. The map of Roueché illustrates well the great irregularity of the portico in all but its central section, with the western part being closely spaced, and the eastern part widely spaced.
Of shops, 4 rooms were excavated towards the west end of the portico, alongside the north-east corner of the basilica, in 1970–71. They face onto this portico and measure internally 4 m by 4 m, according to Stinson (2016) 82. They appear on p. 76 fig. 38, but there is no indication that they are late antique. They could well represent Early Imperial structures. The plan shows that they are secondary to the basilica (as the terminal unit uses the basilica as its west wall), and that they are composed of one single-room unit and one three-room unit (with two rooms having doorways out onto the portico). Their south wall, on the opposite side of the portico, is strongly buttressed in its original construction, suggesting they were two storeys high. At the east end of the portico the solid back wall of the theatre prevents any shops being established here. Over most of the portico, however, the rear wall has not been revealed.
Of paving, the plan of Stinson (2016) 84 fig. 43 shows a small area of the interior of the portico, excavated close to the civil basilica, an adjacent shop, and the portico’s junction with a north-south street, along the west side of the basilica, which once led into the agora. This plan shows a paving in stone slabs. These look reused, on account of their very varied sizes and because they have sometimes been cut-to-fit at angles. This detail suggests that these blocks are stone slabs rather than ceramic tiles, as this kind of modification is more difficult with a tile. The irregularity and large size of some blocks also suggests stone. The slabs are laid in short rows of irregular width in places, but, overall, the pattern is rather chaotic. Sometimes little attempt has been made to make the blocks fit, leading to gaps. Blocks in front of the street junction are larger (ca. 0.4 m by 0.5 m to ca. 0.8 m by 1.15 m, with many rectangular well-fit slabs), whilst those in front of the adjacent shop and basilica, to its east, are smaller (ca. 0.15 m by 0.15 m to ca. 0.8 m by 1 m, with many non-rectangular pieces), suggesting differences in phasing. In fact, the large slabs coincide with the point where a north-south road on the west side of the civil basilica came into the agora, and so likely represent road paving, not portico paving. In contrast, the smaller slabs (underlined above) are only found in front of the shop and the basilica, suggesting that they represent the real late floor of the south portico. These dimensions were obtained by hand-measuring off the plan.
Of dating, an architrave inscription (ALA 66) was found towards the western end of the portico, which records that ‘Philip, son of Herodian, admirandissimus, returning thanks to his own fatherland, covered (i.e. roofed) the two sections’. The rank used of admirandissimus was introduced in the 4th c., but fell rapidly in prestige of use, as ALA commentary V.42 describes, being used for the titulature of a Praetorian Prefect of the East in 368 and for a civic official in 609, passing by the titulature of a πρωτεύων [civic notable] in 430/31 and of an agens in rebus / defensor civitatis in 459. The title is most widespread in the 5th c. Given that Philip does not mention his office, Roueché believes he must be an eminent local citizen or minor imperial official. Roueché thought that the inscription should date from the later 5th to 6th c. It is important to disassociate this inscription from the sequence of acclamatory inscriptions honouring Albinus on the columns of the adjacent west portico of the agora, which need not be of the same date. Unfortunately, Stinson does not do this, as he dates the Albinus inscriptions collectively with the Philip inscription to the late 5th to late 6th c., mistakenly following Roueché. Of the Albinus inscriptions, Roueché has recently noted that the overall dating of such acclamatory inscriptions seems to fall in the time of Justinian up to the early 7th c., thus modifying the range of the Albinus texts: C. Roueché, “From Aphrodisias to Stauropolis”, in Wolf Liebeschuetz Reflected, edd. R. Salway and J. Drinkwater (London 2007) (183–192) 186. However, she says nothing about the Philip inscription, which was not an acclamatory text. Whether the ‘Philip son of Herodian’ inscription dates the whole of the portico is uncertain, especially given the differences in construction observed. Nevertheless, it is fair to note that the dating difference between the inscriptions is not great. Overall, in the absence of other evidence, this Philip inscription gives the best chronological indication for the construction of the late south portico, and so I will use it here, alongside the general presence of reused material, which would give us a generic late antique date. We should disassociate ALA 30 and 198, found on columns on the east end of the portico, as these seem likely to relate to an earlier restoration: see appendix K4a.
Of relationships to other buildings, Stinson notes a piece of sculpted frieze block from the civil basilica, which was incorporated in the stylobate of the portico. He believes this might have come from the attic storey of the north façade, or the vestibule’s upper storey, or the long hall’s upper storey. He speculates that the portico was built across the façade of the basilica to hide its ruined nature: Stinson (2016) 85. However, I think this is an over-interpretation, and I do not accept that he has any evidence that puts the rebuilding of the portico after the abandonment of the basilica. Here we are not dealing with a large amount of spolia from the building, but with one block, and the back wall of the portico is very definitely the basilica face. Furthermore, there is no evidence of blocking of the basilica doors. Rather I think we should see the reused block as debris from a repair to the basilica, likely associated with the installation of the portico. The aesthetics of building this portico across the basilica are not odd from a late antique perspective, which sees many aesthetic disjunctures in the construction of porticoes, sidewalks and shops. Rather, one should think of the amenity the portico provided in linking all sides of the agora, keeping pedestrians dry and shaded. Thus, the portico and the civil basilica could well have coexisted.
Dating summary (for one or possibly two late antique rebuilds): range 450–600, midpoint 525, class Cs5 (absolute, inscription in situ), Cs4 (reused material, general presence), publication 3/3.
13ASI Sagalassos (Lower Agora): The northern half of the east portico was replanned sometime during 450–575, in works that seem to have involved the rebuilding of at least the southern part of the colonnade of the same portico. The development of this area is best described in T. Putzeys, Contextual Analysis at Sagalassos: Developing a Methodology for Classical Archaeology (Ph.D. diss., KULeuven 2007) 215–17 with figs. 81–82 (phase plans) and 263–84, a study which relies on archive materials that were available to its author, but are not described in internal reports or published articles.
Of phasing, the redevelopment can be described as follows: a row of Early Imperial cellular shops and a portico were replaced in the northern half of the portico by a completely irregular plan of one group of interconnected rooms and a single surviving cellular unit, with rubble walls and some reused material. The irregular group of rooms was built over the portico space, as well as over the site where the shops had stood: L. Lavan site observations 2005. This replanning provides the best context for the introduction of two pieces of spolia into the colonnade itself, where it survived in its southern half. This material indicates a rebuilding of the portico. These pieces consist of an altar cut down into a column base, plus a milestone of Constantine and his sons. These are shown on the agora paving within sector 2340–75 on the drawing in site notebook LA (92) 52, in association with other elements of the portico. The texts are published as H. Devijver and M. Waelkens, “Roman inscriptions from the fifth campaign at Sagalassos”, in Sagalassos IV. Report on the Survey and Excavation Campaigns of 1994 and 1995, edd. M. Waelkens and J. Poblome (Acta Archaeologica Lovaniensia Monographiae 9) (Leuven 1997) (293–314) no. 6.1 = SEG XLVII 1997 no. 1770.
Of dating, an associative date for the replanning of the portico was initially given as late 4th or early 5th c., but has been revised by Putzeys, who had access to the site archive and subsequent research, suggesting a much later dating, based on new evidence, for which see appendix Y3. Critically, Putzeys noted that, inside the portico, room E3 (p. 263) produced phase 8 ceramics (450/75–550/75) from the foundation fill / floor of the last reorganisation. Of less importance, Putzeys also noted that the occupation material inside the shops, before the roofs collapsed, produced Sagalassos Red Slip Ware phase 9 (in room E4 on pp. 267–68) or phase 8/9 (in room E8 p. 272), plus coins in room E3 (5 bronze of probably 4th–5th c. date, plus 1 of Constans II (641–68) (p. 264)), plus coins in room E7 (mainly 4th– 5th c., but 1 of the early years of Heraclius (611–13), which was the latest coin (p. 271)).
Of subsequent phasing, we can note that the later ‘guard rooms’, added to the northern end of the rebuilt portico buildings, are very different. They contain much spolia, unlike the late replanned walls of the northern part of the eastern portico: site observation L. Lavan 2005–2006. Thus, I think they belong in a later phase. This detail is not noticed by T. Putzeys et al., “Shops and retail in Late Antiquity. A contextual approach to the material evidence from Sagalassos”, in Thinking about Space. The Potential of Surface Survey and Contextual Analysis in the Definition of Space in Roman Times, edd. H. Vanhaverbeke, J. Poblome, F. Vermeulen, M. Waelkens and R. Brulet (Studies in Eastern Mediterranean Archaeology 8) (Turnhout 2008) (161–217) 180–82. Putzeys describes both the portico buildings and the guard rooms as being built using the same technique of mortared rubble and spolia, without considering the level of spolia content, which is much greater in the ‘guard rooms’. These ‘guard rooms’ have also produced dating evidence that is slightly later than the shops: floor levels (layer 5) “containing 6th–7th c. AD material”: Internal Report (2003) 96–98, which T. Putzeys (pers. comm.) clarifies as being Sagalassos Red Slip Ware phase 8 to phase 9.
Overall, Putzeys situated the replanning of the rooms in the early 6th c. AD, after the earthquake of ca. AD 500, which affected the Agora Gate area and the western side of the Agora, for which see appendix S6 on staircases and F7b on the repair of the gate. However, it is wiser to ignore this catch-all theory, as the evidence from the eastern portico is good quality: it is possible that the building sequences on each side of the agora were not exactly contemporary. Rather, I prefer to use a dating range of ca. 450 to 575, based on the full range of the ceramics from the floor of room E3. The absence of phase 9 pottery (AD 550/75–ca. 700 in 2013) in foundation levels, and its presence above, also suggests a construction date before ca. 575, whereas occupation clearly continued into the mid-7th c., at least as late as 641.
Dating summary (portico redevelopment): range 450–575, midpoint 512.5, class Cs9 (contextual ceramics), publication 3/3.
13ASI Sagalassos (Lower Agora): The west portico of the Lower Agora was rebuilt, along with the portico’s rear wall, which corresponds with the terrace wall of the Apollo Klarios temple temenos to the west. This probably took place at the same time as the Agora Gate, when a water supply was installed, stretching down from the Agora Gate along the colonnaded street to its south. The southern half of the west portico, closest to the Agora Gate, shows most signs of rebuilding, as does the southern half of the rear wall / terrace wall. In the north half of this portico an Ionic colonnade of grey limestone columns and Attic bases is set on two steps of white limestone, whereas on the south half of this portico the (quarry-ready) bases are irregular in style, with an assortment of non-matching columns of different types. The steps of the southernmost part of the portico have been replaced with some pink limestone in the section of the portico where the rear wall has also been rebuilt: L. Lavan site observations, July 2005. L. Loots, The Building Materials and Building Techniques at Sagalassos, Turkey (Ph.D. diss. KULeuven 2001) 415; M. Waelkens et al., “The 1996 and 1997 excavation seasons at Sagalassos”, in Sagalassos, vol. 5, Report on the Survey and Excavation Campaigns of 1996 and 1997, edd. M. Waelkens and L. Loots (Acta Archaeologica Lovaniensia Monographiae 11/B) (Louvain 2000) (217–398) 368 (noting a reused statue base incorporated in the wall); Internal Report (1999) 135–36; Internal Report (2000) 139–45. For other reused statue bases in the wall, see L. Lavan, “The agorai of Sagalassos in Late Antiquity: an interpretive study”, in Field Methods and Post-Excavation Techniques in Late Antique Archaeology, edd. L. Lavan and M. Mulryan (Late Antique Archaeology 9) (Leiden-Boston 2013) (289–353) 295–96, 301–302 with fig. 3 and table 1, 313.
Of phasing, the portico and rear wall form a contiguous area of redevelopment, which partly depends on the retaining of earth by the late staircase of the Agora Gate and contains some of the same spolia types. For example, herringbone-marked blocks from the rear wall of the portico are found in the colonnaded street paving laid over the trench for the new water supply that runs through the Agora Gate after its rebuild, and probably also along the front of the west portico to the nymphaeum: L. Lavan site observations, July 2005 and July 2006. It is likely that the whole area of rebuilding is the result of a collapse of the south-east side of the Apollo Klarios platform in the earth movements that resulted in the collapse and the rebuilding of the Agora Gate.
Of dating, chronological evidence for this phase of building work comes from the staircase of the Agora Gate area after the ‘earthquake of AD 500’ [no Guidoboni Catalogo entry]. This phase of construction is dated by a fill supporting the new water supply, set as part of the same works against the stairway of the Agora Gate, which incorporated pottery dated to the “first half of the 6th century AD” and coins dating from the 3rd to the 5th c. AD (the latest two being “4th to 5th c. AD”), for which see appendix S6 on staircases and F7b on the repair of the gate. I will use the whole period cited for the pottery to provide a contextual date for the works, considering the portico to be part of the same phase of development.
Dating summary: range 500–50, midpoint 525, class Cs3 (associative, phase of development), Cs9 (contextual pottery), Cs8 (contextual coins), publication 2/3.
See also (appendix K4a): 15ORI Antioch in Syria (Hellenistic Agora): ‘Basilica of Rufinus’ (likely a double-aisled portico) rebuilt in 507–18.
See also (appendix S1): 13ASI Laodicea ad Lycum (North Agora) (3): Three porticoes were built as part of the works, with opus sectile in one portico 494–610.
See also (appendix S3): 06ITS Terracina: The forum was cleaned up in 554–661, probably involving the rebuilding of a portico, as an inscription recording works is inscribed on a column; 12CPL Constantinople (Forum of Theodosius) (2): Two double-aisled porticoes were rebuilt, one faced with piers; 12CPL Constantinople (Basilica) (4): 4 surrounding porticoes were likely rebuilt as part of Justinianic works, 532–37; 12CPL Constantinople (Augusteion): Surrounding porticoes were likely rebuilt as part of Justinianic works, 532–44; 15ORI Scythopolis (3): New portico built and others refloored with black and white tiles as part of early 6th c. restoration of the plaza, 500–535.
See also (appendix Y2) 13ASI Sagalassos: The west portico on the Upper Agora saw building work that reused some architectural elements, in a phase that can be ascribed to the date range 450–575 based on ceramics recorded. This phase entirely obliterated / covered any portico that had existed here previously. The portico measured ca. 5.8 m deep, was single-storey, and included reused Doric columns of different types / possible weathered Ionic columns, and pedestals, for which they were sometimes a bad fit, fronting a row of shops, of which short shafts survived, some with concave faceting, some without, some upside down, placed on pedestals with integrated column bases, which look like a variant on Ionic. The rear of the northern part of the portico was covered in marble veneer.
S5a Paving / Surfacing
02HIS Ilici: There was a ‘levelling’ of the forum area, with spoliation ditches filled in, perhaps to make a new ‘plaza’ by the temple (or at least an open space). The temple was transformed into a church at the same moment [apparently because of the relationship of the levelling layer with the new entrance]. As part of this work, the temple, which was oriented roughly east-west, saw the removal of its façade in antis. It also saw the division of the cella into two parts, with a new wall that may have supported a chancel screen, fragments of which were found in the structure: J. Molina Vidal and A. Poveda Navarro, “El nivel de abandono de un sector del foro de Ilici”, Actas del XXIII Congresso National di Arqueologia (Elche 1995) (Elche 1997) (141–54) 150–52. I was not able to obtain A. Poveda Navarro, “Aproximación al urbanismo de Ilici Augusta durante la Antigüedad tardía”, in VI Reunió d’Arqueologia Cristiana Hispànica: les ciutats tardoantigues d’Hispania: cristianització i topografía. València, 8, 9 i 10 de maig de 2003, edd. J.M. Gurt Esparraguera and A. Ribera i Lacomba (Valencia 2005) 323–44 (not seen). For this reason, the page nos. that follow are derived from the 1997 article.
Of dating, we have two elements. (i) Firstly, the fill of the levelling layer over the forum has been dated to the 7th c.: see ceramic evidence of UE 8 on p. 150. This is based on the latest ceramic material, which are cooking wares “IIA” (1 sherd) and “I2D” (5 pieces) and coarse ceramics (10 pieces), all dated to the last quarter of the 6th c. to the first quarter of the 8th c., although it is believed that the coarseware might possibly go later. Earlier material stops at AD 500, at the very latest, and forms two distinct groups, the first of 400 BC to AD 25 and the second of AD 75 to 500, with two ceramics overlapping. This suggests that the earlier pieces are residual and that the fill dates from the time that the later pottery was deposited, so sometime in AD 575–725. Given that this is a fill that was likely brought from elsewhere, not occupation debris, we should place the construction event of the levelling of the plaza anytime up to 725, the end date of the cooking wares. (ii) Secondly, we can also note that the chancel screen was identical to those found in a nearby Christian basilica and dated to the 7th c.: Molina Vidal and Poveda Navarro (1997) 150–52. The work of A. Ramos Folqués, “Un cancel visigodo en La Alcudia de Elche”, Pyrenae 8 (1972) 167–72, dates the chancel that is being used as a parallel here, but mainly by reference to the creation of a bishopric in Ilici, which can be detected for the first time in 630–42 at the 4th council of Toledo, thanks to the signatures. This creation of a bishopric is thought to have triggered the creation of the mosaic-encrusted basilica of the city. As this is not very strong evidence, I will rely on the ceramics alone, to suggest a date corresponding with the entire range of the final identified pottery, although this is not a true contextual date, as we cannot see a range of differently dated material here, only a single date range. It is more of an associative date, taking the last major finds to indicate the latest activity.
Dating summary: range 575–725, midpoint 650, class Cs9 (contextual ceramics), class Cs3 (associative, phase of development), publication 3/3.
06ITS Rome (Forum Pacis): The buildings on the plaza were levelled in the 6th c., creating an open space again, in a part of the plaza that had become occupied. Structures built within the square were filled with masonry and with a level of sandy earth; the site was again made into an open plaza, before subsequently becoming a cemetery: see preliminary report, without dating details, of R. Santangeli Valenzani, “I fori imperiali nel Medioevo”, RM 108 (2001) (269–83) 269–71.
Dating summary: range 500–600, midpoint 550, class 0, publication 1/3.
13ASI Xanthos (Lycia) (Upper Square): The agora pavement seems to have been renewed, although very few details are available. Sherds found under the pavement (from sondages undertaken in gaps where the slabs were missing) are dated to the 5th–7th c.: L. Cavalier, pers. comm. These included coarsewares with painted bands, as well as LRC (Phocaean Red Slip Ware) and LRD (Cypriot Red Slip Ware) of 5th–6th c. date: C. Rocheron pers. comm. Without precise knowledge of the wares, it is difficult to give this development anything other than a TPQ of the 5th c. and a TAQ of 614, the date of the decisive Persian invasion of Asia Minor, after which no civic secular public building is known. Following the expulsion of the French Mission from Xanthos it is now unlikely that the fine ware types will ever be identified.
Dating summary: range 400–614, midpoint 507, class Cs7 (TPQ ceramics), z (regional development), publication 3/3.
13ASI Ephesus (Tetragonal Agora): There was apparently a marble floor laid on the agora in its 6th c. phase: see P. Scherrer, “The stratigraphy of the tetragonos agora”, in The Amphorae of Roman Ephesus, T. Bezeczky (Forschungen in Ephesos 15/1) (Vienna 2013) (5–17) 16–17. This analysis mentions an amphora type (phase 8 no. 298) coming from layers and pits beneath the latest marble floor in the agora (laid down in the mid-6th c. AD). However, on p. 121 it is clear that this amphora is Dressel 6A, given as dating not later than the 1st c. AD. There are no other amphorae in layers of this phase. However, p. 17 (table 2.1.4) shows that the subsequent disturbed phases have produced material up to the early 7th c. AD, with many amphorae listed, most clearly by type on pp. 265–67, such as the Samos Cistern Type, given as of 500–700 on RADR. Thus, a TPQ is not provided by the Dressel amphora, which is too early to provide a meaningful contribution. The basis for the layer date given in the amphora report probably only comes from the wider late phase of development in the agora of this same time. We are not told that the new floor matches the 6th c. higher level seen in the north portico, so I will have to assume this, pending further publication, and adopt a date based on phase of development. Thus, we must adopt the chronological framework of the 6th c. north portico and its associated developments. This has a dating range of 550–610, based on the structural integration of the late agora repairs with the city wall, and on the presence of types of spolia also present in neighbouring building work dated to the 6th c. (see appendices C3 (for relationship to Arcadiane and late city wall), K2b (for spolia around Lower / Tetragonal Agora) and S3 (for 6th c. rebuild of agora)). A generic regional TAQ is provided by the nominal end of secular civic monumental building after the Persian invasion of 614, as elsewhere in Asia Minor. The date range generated is poor, as the reasoning that associates the paving to the wider dating phase is not made clear from the reports.
Dating summary: range 550–614, midpoint 582, Cs3 (associative, phase of development), z (regional development), publication 1/3. Poor.
13ASI Sagalassos (Lower Agora, paving by Agora Gate): The southern section of the agora seems to have been repaved, along with the rebuilding of the Agora Gate. The paving here has a strip of very mixed stone types (containing reused material) in contrast to the rest of the stone types in this paving. This was revealed by P. Degryse, in work undertaken by him for me in 2006, with my spolia observations on site in 2005–2006. Our remarks relate to two areas. (i) The last 5 narrow rows of slabs on the southern side of the plaza cover an area of about 19 m by 3 m. This area is composed of rectangular blocks sorted by size, then laid in parallel rows of irregular width, similar to the main paving of the agora, to the north. However, their stone types are far more heterogeneous, if not obviously reused, than the paving to the north. Furthermore, the rows are more uniformly narrow than those of the slabs to the north. There seems to be a higher preponderance of mason’s marks on these slabs than elsewhere in the agora. The slabs measure ca. 0.4 m to 0.6 m in their width, whilst the length of the individual slabs within the rows are from 0.3 m to 1.2 m. (ii) A further 1–3 chaotic rows of larger slabs borders the 5 narrow rows to the south, covering an area of around 17 m by 1.5 m to 2.5 m. This area (of 1–3 chaotic rows) is composed of obviously reused rectangular slabs, including examples that fit badly, and were not re-cut, with some surfaces showing recessed portions for jointing with other blocks. They are laid in approximative rows, without any attempt to cut corners to fit. Their sizes are from ca. 0.4 m by 0.6 m up to ca. 1.5 m by 1.7 m. Mason’s marks do not seem to be present.
Of phasing, we have to set the paving within a wider understanding of the Lower Agora. It seems likely that the west portico and the southern part of the west portico rear wall (retaining the Apollo Klarios terrace) were rebuilt at the same time, as the area of rebuild corresponds exactly. This rebuilding forms part a larger contiguous area of rebuilding in reused material around the Agora Gate, after it collapsed in a trauma that probably saw the Apollo Klarios terrace disintegrate. The rebuilding work, which joins up all this area, includes the installation of a water supply that extends down the colonnaded street to the west. This area also shares reused stones of similar types. Thus, we can suggest that all this area of repair work dates to the same time as the remodelling of the adjacent Agora Gate, after the ‘earthquake of AD 500’, for which see appendix S6 on staircases and F7b on the repair of the gate. Given that the Agora Gate staircase retains earth that supports the agora slabs, we can envisage that at least the last 3 chaotic rows, which adjoin it, were laid as part of this work. There is a chance that the ‘5 neat rows’ date from an earlier phase, but one could envisage that this area was relaid, whilst the ‘chaotic area’ was replaced with new materials.
Of dating, chronological evidence for this phase of building work comes from the staircase of the Agora Gate area after the ‘earthquake of AD 500’. This phase of construction is dated by a fill supporting the new water supply, set as part of the same works against the stairway of the Agora Gate, which incorporated pottery dated to the “first half of the 6th century AD” and coins dating from the 3rd to the 5th c. AD (the latest two being “4th to 5th c. AD”), for which see appendix S6 on staircases and F7b on the repair of the gate. I will use the whole period cited for the pottery to provide a contextual date for the works, considering the portico to be part of the same phase of development. We also have a TAQ for the paving: a 7th c. water supply crosses over the chaotic paving area on the site drawings from SA92 LA, in the Sagalassos site archive, but this detail has not been carried over into other official site drawings, such as T. Putzeys, Contextual Analysis at Sagalassos: Developing a Methodology for Classical Archaeology (Ph.D. diss., KULeuven 2007) 214 fig. 79.
Dating summary: range 500–50, midpoint 525, class Cs9 (contextual pottery), Cs8 (contextual coins), publication 2/3.
See also (appendix S1): 15ORI Bostra (Bosra) (rectangular plaza): Rectangular slabs were laid in rough rows of irregular width, sometimes interrupted, with no examples of angles cut to make different slabs fit. The presence of slabs of very varied sizes in front of the south portico makes it possible that some of them were reused. The size of the slabs varies between ca. 0.4 m square and ca. 1 m square. The dating is of 487.5–512.5. 16AEG Abu Mina (Pilgrims’ Court): The square seems to have been covered in what looks like new-cut stone slabs sorted into straight rows of uneven width, arranged in more than one system of alignment. Slab sizes vary from ca. 0.42 m by 0.48 m to ca. 2.4 m by 1.1 m. Dating is 475–532.
See also (appendix S2): 10MAC Dyrrachium: The round plaza was paved in grey-white rectangular marble slabs, 0.6 m thick and apparently newly cut, with no signs of reuse. The slabs, measuring between ca. 0.5 and 3.4 m in length and similar varied widths, were laid in rows of different widths in a concentric radial arrangement, tapered to fit, except for an outer row of larger slabs, 0.82–0.9 m wide, which was laid lengthways. The plaza is dated to 500–25. 09DAC Justiniana Prima: The plaza was covered with stone slabs, more or less the same size, of no more than 1 m square, irregular in form, not especially rectangular. Dating is to 535–616.
See also (appendix S3): 12CPL Constantinople (‘Basilica’ courtyard): The paving was of large Proconnesian marble slabs of different lengths and widths, from 0.7 m to 1.05 m by 1.5 m to 2 m. Dating is of 532–37.
See also (appendix V4a): 06AFR At Carthage, just prior to the reconquest, buildings on the Admiralty Island plaza were levelled, although graves seem to have been cut within the square around the same time [500–525].
Discounted: 06ITS Florence: Undated repairs to the paving of the forum have been detected: R. Mirandola, “Firenze”, in Archeologia urbana in Toscana, ed. S. Gelcihi (Mantua 1999) (59–72) 67, who mentions repairs using bricks; R. Davidsohn, Storia di Firenze, vol. 1 (1956) 78 n. 3, citing a description of the find in Le Nazione no. 393 of 29th October 1892 (not seen), mentions patching (“fu rappezzato”) of marble with bricks, and puts this in the Forum Baths, rather than the Forum itself; C. Corinti, “Firenze antica nei disegni di Corinto Corinti”, L’universo 56 (1976) (1071–1143) map 67 (not seen). This work was supposedly the result of an intervention by Narses, based on the idea that he restored Italian cities after the Gothic Wars, as recorded in Marius of Aventicum, Chronicle AD 568 (MGH.AA, vol. 11 p. 238). However, we cannot date this repair archaeologically. Furthermore, Florence of course also saw medieval urban occupation, so this work could have been done in much later centuries.
S5b Arches and Entrances
See also (appendix F7b): 13ASI Sagalassos (Agora Gate): The ruined Agora Gate, where the north-south colonnaded street joined the Lower Agora, was partially re-erected, to exhibit some relocated statuary, as attested by a non-original lion capital (which may have carried a statue) and the base for an infant statue, set, secondarily and rather oddly, on the west podium. Dating is to 500–550. 13ASI Sagalassos (Arch on the road leaving the Upper Agora): The structure was patched up and retained, within a comprehensive replanning of the street, in which heavy walls of spolia blocks were erected to form a passage around the street on each side of the arch. These works sought to hide damage, and used displaced wall-moulding from the arch as a decorative feature of the new passage walls. Dating is to 475–525.
See also (appendix F8a): 15ORI Caesarea Palestinae (Caesarea Maritima) (Wall arch on Cardo E2.): This triple-portalled arch was built as an entrance into the Byzantine esplanade, which was dedicated in a mosaic inscription as “the arch (ἁψίς), together with the wall and the staircase from city funds”. The text informs us that this occurred during the mandate of the general and proconsul Flavios Entolios, sometime between AD 546 and 614. 16AEG Abu Mina: There is a wall-arch at the southern end of the southern half of north-south colonnaded street, at its junction with the Pilgrim’s Court. The colonnaded three-portalled wall-arch is just one column-width thick, with a Corinthian capital. Dating is of 475–619.
See also (appendix K2b): 13ASI Laodicea ad Lycum (North Agora): Propylon 1 saw some rebuilding for a second time in 494–610, although the architectural form of the rebuild is not yet clear, the rest of the plaza being significantly transformed after the earthquake of 494.
See also (appendix S1): 15ORI Bostra (Bosra) (rectangular plaza): A monumental entrance cut through a row of shops, to create an access from the street into the plaza. The central portal measures 3.4 m wide, flanked by two side-portals 2.4 m wide, in a wall that replaced the back wall of the shops and was slightly thicker than it (so the arch was a thin arch, rather than an arch with piers). All three portals were likely covered by arches. The entrance may have been fronted by columns, to create a monumental vestibule, on the face onto the street or that onto the plaza. The dating is of 487.5–512.5.
See also (appendix S2): 09DAC Justiniana Prima (4): It is possible that 4 arches covered the entrances into this round plaza from the 4 streets that intersected there, as revealed by thicker piers at some of the junction points. The dating is of 535–616.
See also (appendix V4b): 07AFR Carthage (Admiralty Island plaza / ‘Maritime Agora’): 4 poorly-aligned piers, perhaps representing a tetrastylon or tetrakionion, were erected to replace the monumental entrance of the plaza in the first late phase of renewal, dated from 500 to 525.
S6 Staircases and Ramps
13ASI Ephesus (Tetragonal Agora): Ramps [I do not know how many but suspect two] were added to the Lower Agora, containing much reused material. This can be dated in one case: S. Karwiese, “Das Südtor der Tetragonos Agora in Ephesos”, ÖJh 66 (1997) 306, 315, sealing a coin of AD 544/45, with L. Lavan site observation 2005. It is unlikely that the works, which do not greatly alter the traditional character of the agora, date to after the decisive Persian invasion of 614, after which secular civic public building work is not known in Asia Minor. Coins from the adjacent areas south of the city wall extend as far as Heraclius (610–41), but not as far as those of his successor Constans II (relatively common in the region), suggesting that much of the area was given up during the reign of Heraclius: see appendix H7 (preliminary comments on the statue monuments of the Embolos).
Dating summary: range 544–614, midpoint 579, class Cs7 (TPQ coin), z (regional development), publication 3/3.
13ASI Sagalassos (Agora Gate): The Agora Gate saw a destruction layer of the early 6th c. cover the original staircase, whereby the north-south colonnaded street entered into the Lower Agora. Subsequently there was a relaying of a new staircase, made in part from reused fragments of the Agora Gate. This staircase was ca. 10 m wide and ca. 23 m long, including two flights of stairs, one ca. 2 m long at the top and the second ca. 7.5 m long at the bottom, measuring off the plan cited below [and based on L. Lavan site observations 2006]. The staircase extended across the width of the former roadway, except for a strip of the road of 3.5 m to 4.5 m, which contained a new water supply. The fill [layer 6] supporting this new water supply contained pottery of the “first half of the 6th century AD” and coins of the 3rd to the 5th c. AD (the latest two being “4th to 5th c. AD”). The two elements (i.e. the new staircase and the rearranged water supply) form a single unified phase of the replanning of the road. A second, early 7th c. destruction layer, lies on top of the 6th c. AD stairway: M. Waelkens et al., “The 1994 and 1995 excavation seasons at Sagalassos”, in Sagalassos IV. Report on the Survey and Excavation Campaigns of 1994 and 1995, edd. M. Waelkens and J. Poblome (Acta Archaeologica Lovaniensia Monographiae 9) (Leuven 1997) (103–216) 208–10 with 207 fig. 165 (plan) and 211 to 213 (figs. 173–74). I will use the whole period cited for the pottery to provide a contextual date for the works, both the water supply and the stairs associated with it in the same phase of works. See also appendix F7b on the repair of the gate. The white limestone staircase is surprisingly worn, towards its centre, despite being a 6th c. creation: L. Lavan site observation, July 2005. This attests to the high volume of pedestrian traffic on the street.
Dating summary (for new staircase): range 500–50, midpoint 525, class Cs9 (contextual ceramics), Cs8 (contextual coins), publication 2/3.
13ASI Sagalassos (Lower Agora): A short staircase, leading north from the Lower Agora, probably dates from the same time as rebuilding work on the west portico, as the two structures incorporate similar types of reused material (pink limestone): L. Lavan site observation 2005. The west portico rebuild is itself dated by the associated rebuilding of the adjacent Agora Gate, where a fill context supporting a water supply installed during the works has produced pottery dated to the “first half of the 6th century AD” and coins dating from the 3rd to the 5th c. AD (the latest two being “4th to 5th c. AD”), for which see appendix F7b on the repair of the gate. I will use the whole period cited for the pottery to provide a contextual date for the works, considering the portico to be part of the same phase of development. Slightly less precise but supporting evidence comes from the street running north from the short staircase. Here there was no paving, but final surfaces / substrata contained, as their last ceramics, wares from AD 450–75 and AD 550–75, and from the mid-5th to mid-6th c.: F. Martens, “Late antique urban streets at Sagalassos”, in Technology in Transition A.D. 300–650, edd. L. Lavan, E. Zanini and A. Sarantis (Late Antique Archaeology 4) (Leiden 2007) (321–65) 353, 355, drawing on F. Martens, Interdisciplinary Research Concerning the Urban Development of Sagalassos. Settlement Development, Urban Layout and Infrastructure (Ph.D. diss., KULeuven) 450–51; M. Waelkens et al., “Report on the 2002 excavation and restoration campaign at Sagalassos”, XXV Kazı Sonuçları Toplantısı 1.Cilt. (2004) (215–30) 221, and Sagalassos Internal Report (2002) 109–13. Overall, it seems likely that a plan of redevelopment was carried out on the whole north-south street, in the first half of the 6th c., as discussed in appendix C3, of which this staircase is part.
Dating summary: range 500–50, midpoint 525, class Cs3 (associative, phase of development), Cs5 (catch-all masonry, specific spolia) Cs9 (contextual pottery), Cs8 (contextual coins), publication 2/3.
13ASI Sagalassos (Upper Agora): A monumental staircase of white limestone was subsequently constructed over the portico and the shops, built on the west side of the agora: Sagalassos Internal Report (1999) 19–20 with p. 9 fig. 3 for a good plan; Sagalassos Internal Report (2000) 10; T. Putzeys Contextual Analysis at Sagalassos: Developing a Methodology for Classical Archaeology (Ph.D. diss., KULeuven 2007) 291–99, esp. 292. It had a width of ca. 6.8 m and a length of ca. 8 m, hand-measuring off 286 fig. 108 (not easy to estimate, as the staircase is ruinous on this plan). It led up to the former bouleuterion. This bouleuterion had now lost its roof and become an atrium courtyard, whilst its own former courtyard had now become a church, entered from the atrium. The dating of the staircase is explained in appendix Y2. Overall, we can give the staircase a TPQ of after 450 based on the ceramics from the portico foundation, and an associative, phase of development date of the full range of ceramics of 525–575 coming from the construction of the church that it served. The latter dating is given priority. Sagalassos reports do not list individual wares, which might permit tighter dating, so we must use this full dating range.
Dating summary (phase 2: stairs to church): range 525–75, midpoint 550, class Cs9 (contextual pottery), Cs3 (associative, phase of development) publication 2/3.
15ORI Hippos / Sussita: On the north side of the forum, between the main forum and the ‘Hellenistic compound’ to its north, an area of pavement ca. 4 m wide, made of reused blocks, and a staircase, seem to have been built (of which the foundation of reused blocks survives, with two steps in one place). Of dating, for these structures, the report tells us that “the pottery, building materials, and construction methods all testify that both the pavement and the staircase were built during the Byzantine period”: A. Segal et al., Hippos-Sussita. Fourth Season of Excavations June–July 2003 (Haifa 2003) 5; A. Segal et al., Fifth Season of Excavations (September–October 2004) and Summary of All Five Seasons (2000–2004) (Haifa 2004) 17–18, 30–31 with figs. 7–8 for plans. From plan 7, I hand-measured a maximum possible width for the staircase of 44.5 m, although it is not clear if the staircase extended along the full width of the north side of the forum, partly because it is not fully excavated. I have taken ‘Byzantine’ to mean 400–636, as it often is considered to be in archaeological reports on sites in Israel. Thus, the different levels of the forum and the Hellenistic compound were joined together.
Dating summary (establishment of staircase): range 400–635, midpoint 517.5, class Cs9 (contextual ceramics), publication 2/3. Generic late antique date.
See also (appendix S3): 15ORI Scythopolis (‘Byzantine Agora’): An entrance ramp into the trapezoidal square bears an inscription of the governor Romytalkes and is part of works dating from 500–35.
S7a Monumental Fountains, Repairs
13ASI Sagalassos (Severan Nymphaeum, Lower Agora): Repairs in non-matching reused material are visible on the west side of the nymphaeum, to the penultimate projecting podium. The crowning block of the podium is reused (it has non-matching moulding), as is the column base set on top of it (which is quarry-ready rather than Ionic): L. Lavan site observations, July 2006. It is tempting to think that these were part of an earlier building, taking material from an earlier Trajanic nymphaeum on the same site: L. Loots, The Building Materials and Building Techniques at Sagalassos, Turkey (Ph.D. diss. KULeuven 2001) 112. However, no definite origin has been so far been established for these blocks. It is likely, however, that this work is contemporary with the repair of the adjacent agora west portico and rear wall. These were rebuilt as part of a contiguous area of building, following the earthquake of ca. ‘AD 500’, which is dated by ceramics of the first half of the 6th c. and by “4th–5th c.” coins, both coming from a fill supporting a water supply, associated with the relaying of the Agora Gate stairs. I will use the whole period cited for the pottery to provide a contextual date for the works, both the water supply and the stairs associated with it in the same phase of works. In short, the work on the nymphaeum seems likely to be the northernmost tip of a contiguous area of rebuilding, which extends along the entire west portico as far as the Agora Gate: L. Lavan site observations, plus see details in appendices F7b, S4 and S6, with details of the earthquake. The establishment of a new drain for the nymphaeum, cut somewhat crudely into the damaged agora paving, to run in front of the west portico, but never finished, probably dates from the same period, as its line runs directly towards the Agora Gate: L. Lavan, “The agorai of Sagalassos in Late Antiquity: an interpretive study”, in Field Methods and Post-Excavation Techniques, in Late Antique Archaeology, edd. L. Lavan and M. Mulryan (Late Antique Archaeology 9) (Leiden-Boston 2013) (289–353) 306–308 with fig. 5.
Dating summary: range 500–50, midpoint 525, class Cs3 (associative, phase of development), Cs9 (contextual ceramics), Cs8 (contextual coins), publication 2/3.
13ASI Sagalassos (Antonine Nymphaeum, Upper Agora) (2): Poorly refitted blocks in the rear façade wall, indicating a rebuilding in the original material, are noted in M. Waelkens et al., “The 1994 and 1995 excavation seasons at Sagalassos”, in Sagalassos IV. Report on the Survey and Excavation Campaigns of 1994 and 1995, edd. M. Waelkens and J. Poblome (Acta Archaeologica Lovaniensia Monographiae 9) (Leuven 1997) (103–216) 147–62 and in conservation reports in Internal Reports from 2001 onwards. These poorly refitted blocks were revealed by preparation of the stone for the anastylosis, which has restored the monument to its late antique condition. On site, I also noted two quarry-ready column bases, not matching the column diameters in the west aedicula, and one on the west side of the central niche: L. Lavan site observation 2006. However, these stones were placed here using blocks found inside the structure as part of its anastylosis: site plans, available to me at the time of my work in Sagalassos in 2005–2007, showed the podium without the bases in situ. I am not yet able to confirm the scientific basis of the anastylosis details, which is likely to relate to the position of clamp marks etc. A summary is F. Martens, Interdisciplinary Research Concerning the Urban Development of Sagalassos in Pisidia. Settlement Development, Urban Layout and Infrastructure (Ph.D. diss., KULeuven 2004) 541–43.
At the same time, or sometime afterwards, non-original religious statues were introduced, alongside poorly fitting Early Imperial civic honorific monuments: see appendix H3. For a map of the statue base finds, see L. Lavan, “The agorai of Sagalassos in Late Antiquity: an interpretive study”, in Field Methods and Post-Excavation Techniques in Late Antique Archaeology, edd. L. Lavan and M. Mulryan (Late Antique Archaeology 9) (Leiden-Boston 2013) (289–353) 321–23 with fig. 10b. An unreadable Christian inscription was painted on the archivolt of the central niche of the fountain, in red paint: Waelkens et al. (1997) 161. At some point, the penis was removed from three of the statues (at least two of which were part of the original decor of the monument), which, in the context of the removal of male genitalia elsewhere in Asia Minor, is likely to be deliberate: see appendix H3.
However, a further later monumental phase can be detected in the structure: one statue of Dionysus bears traces of mortar, which suggests that it was walled up at some point, probably with the andesite blocks found concentrated around the lateral aediculae in which the Dionysus statues were housed. Some of these blocks were decorated with crosses: Waelkens et al. (1997) 161. This act of ‘hiding’ mythological sculpture, concealing a figure that was widely celebrated in late antique culture, reflects a very different mentality. It suggests a very different rationale to that which involved the rebuilding of the structure and its decoration with imported mythological statues of the pagan gods, clearly recognisable, though shorn of their penises. Thus, two late phases of building work are likely, with the andesite blocks bearing crosses being added later, perhaps in the later 6th c or early 7th c. However, we cannot distinguish between the dating of these two phases archaeologically.
The rebuilding of the monument is often assumed by the excavators to date to after the earthquake of ca. AD 500 [not in Guidoboni Catalogo]: M. Waelkens et al., “The late antique to Early Byzantine city in south-west Anatolia. Sagalassos and its territory: a case study”, in Die Stadt in der Spätantike—Niedergang oder Wandel? Akten des internationalen Kolloquiums in München am 30. und 31. Mai 2003, edd. J-U. Krause and C. Witschel (Historia Einzelschriften 190) (Stuttgart 2006) (199–255) 228–29. This earthquake is itself dated by ceramics and coins from a deposit contemporary with the re-laying of the (lower) Agora Gate stairs: see appendix F7b. However, there is no unambiguous earthquake damage on the nymphaeum, only evidence of a rebuild. No dating has yet been solidly advanced for these late works. The aesthetic disjuncture, visible in the poorly refitted decoration and in the use of non-fitting reused statues, suggests a date after the mid-3rd c. The honorific statues found within the structure all date prior to this time, according to N. Vanderhulst, The Inscriptions from the City of Sagalassos: a Topographical Analysis including a Text Edition (M.A. diss., KULeuven 2003), using numbering illustrated on Lavan (2013) 321–23 with fig. 10b. The removal of the penises suggests a date after 400, as does the appearance of crosses on public monuments (see foreword to appendix H3, referring to crosses as Christian markers, which are seen on public monuments from ca. 400). All of the works should date before the decisive Persian invasion of Asia Minor in 614, after which secular civic building is not known.
Dating summary (for the rebuilding / installation of non-original statues / penis removal): range 400–614, midpoint 507, class Cs4 (reused material), Cs5 (catch-all, aesthetic disjuncture), z (regional development), publication 3/3. Generic late antique date.
Dating summary (the subsequent walling up of the nymphaeum’s lateral aediculae with blocks marked with crosses): range 400–614, midpoint 507, class Cs1 (architectural and artistic style), Cs5 (catch-all, aesthetic disjuncture), z (regional development), publication 3/3.
S7b Fountains and Basins, New
See also (appendix H6): 13ASI Sagalassos: The small fountain house and the related fountain with apsidal basin, both located in front of a piazzetta east of the heroon, by the north-west Gate, are dated to 500–525. Neither feature exceeds 2 m square in size.
See also (appendix S1): 16AEG Abu Mina (Pilgrims’ Court): a fountain with a vaulted canopy was established close to the centre of the square, looking down the north-south colonnaded street, probably within the period 475–532.
13ASI Laodicea ad Lycum: A fountain with a basin measuring 3.15 m by 1.40 m was established on the late Central Agora within the period in which its construction occurred, although not necessarily in the same phase, within 494–610.
See also (appendix Y2): 13ASI Sagalassos: a long basin was added to the front of the west portico in 525–75, in association with piping and industrial activities added in phase 3 of this complex.
S10a Horologia
12CPL Constantinople (‘Basilica’ courtyard): The horologion near the Milion was made by Justinian, according to Theophanes A.M. 6028 (AD 535/36 date in edn. of Mango and Scott (1997)). However, Malalas states that the horologion was simply moved to be near to the Augusteion and Basilica courtyard at around this time: Malalas 18.85 (AD 538, consulship of John the Cappadocian). The two accounts likely refer to the same event. The location of the Milion and the Augusteion / Basilica is close enough to be the same place. I prefer Malalas’ date, as he was writing at the time of Justinian. These descriptions do not allow us to be sure if the horologion stood directly in front of the Basilica courtyard, but a text recording a repair to a horologion (below) suggests that it was definitely part of the Basilica, unless there were two horologia.
Dating summary: 538, class x (historical text), publication 3/3.
12CPL Constantinople (‘Basilica’ courtyard): Justin II and his wife Sophia (reigned 565 to 574) gave a bronze horologion (which was small enough to steal). This was set in the arch (ἁψῖδα) of the Basilica: Anth. Graec. 9.779, a text which records that it was restored by Julianus the Prefect (PLRE 3.735–36 Julianus 15) after being stolen. The arch might possibly be one of the two structures that initially held statues of the Tyches of Rome and Constantinople (see appendix T2). As Julianus was prefect in 565–66, and Justin’s accession was in 565, this gives little time for the initial gift; the restoration and gift of the horologion were probably the same event.
Dating summary: range 565–74, midpoint 569.5, class x (historical text), publication 3/3.
13ASI Ephesus (Tetragonal Agora): A building surviving as a low opus caementicium foundation faced with ashlar blocks, measuring 10 m by 5.8 m, oriented north-south, was excavated slightly west of the centre of the agora. This structure has produced late 5th c. coins from layers associated with / underlying its construction: P. Scherrer, “Die Agora: Vorläufiger Baugeschichte”, in Forschungen in Ephesos XIII.2 Die Tetragonos Agora, P. Scherrer and E. Trinkl (Vienna 2006) (49–54) 52 with fig. 72 (photo). Whilst it need not be a horologion, the position, as an isolated structure within an agora, still speaks in favour of this, at this date. For the original survey see W. Wilberg and J. Keil, Ephesos III.1. Die Agora (Vienna 1923) 18 with fig. 28 (plan). Although the exact start date for the late 5th c. coins is not known, we can take the fact that several coins were recovered as providing a contextual date of the 25 years following 487.5, which is where late 5th c. begins under the terminology used in this study.
Dating summary: range 487.5–512.5, midpoint 500, class Cs8 (contextual coins), publication 2/3.
13ASI Aphrodisias: A round altar left in place by builders of the tetrastoon was later used as a sundial (ALA 249). It is dated by Roueché in ALA to the 5th–6th c. based on a calibration to the latitude of Alexandria, when the philosopher Asclepiodotus of Alexandria visited the city in the late 470s or 480s (before 489), one of whose entourage may have brought a portable sundial to the city: ALA commentary V.15–16 with references to ancient sources that suggest approximate dating parameters. This sundial has been published as P. Pattenden, “A late sundial at Aphrodisias”, JHS 101 (1981) 101–12. Roueché does not explain her dating, but it is likely based on the presence of the philosopher and wider regional development: there is no recorded civil secular public building in Asia Minor after the decisive Persian invasion of 614. These are both the criteria that I use here.
Dating 470–614, midpoint 542, class x (historical text), z (regional development), publication 3/3.
13ASI Sagalassos: In the small plaza by the north-west gate a reused base and column supporting a sundial, was set over a destruction layer, containing pottery dating to the “first half of the 6th c. AD” and coins of the later 4th and 5th c. AD: Waelkens M. et al. (1997) 191. See appendix H6 (relating to fountains installed here) for further details of the levelling in this area, which is contextually dated based on coins to the first quarter of the 6th c., which permits a closer dating than the ceramics do here. Similar sundial towers occur in 3rd–4th c. Roman sarcophagi, outside of cities, Vienna Genesis illustrations (early 6th c.) and by implication in the Brescia casket (late 4th c.), where the cock crowing at Peter’s denial of Christ is set on a pillar.
Dating summary: range 500–525, midpoint 512.5, class Cs8 (contextual coins), Cs9 (contextual ceramics), Cs3 (associative, phase of development), publication 3/3.
15ORI Gaza: A horologion was found in the heart of the city, before the Basilica Stoa, which sounds like it could be on an agora: Procopius Gazaeus, Horologion. This oration does not mention that it is a new building or that it was restored, but this is possible, as the completion of such works would have provided a natural occasion for delivering such a speech. For commentary see B. Bäbler and A. Schömberg, “Prokop: Die Kunstuhr in Gaza”, in Rose di Gaza. Gli scritti retorico-sofistici e le Epistole di Procopio di Gaza, ed. E. Amato (Hellenica 35) (Alessandria, Italy 2010) 528–59 with earlier bibliography, notably H. Diels, Über die von Prokop beschriebene Kunstuhr von Gaza: mit einem Anhang enthaltend Text und Übersetzung der Ekphrasis hōrologiu des Prokopios von Gaza (Abhandlungen der Preussischen Akademie der Wissenschaften. Jahrg. 1917. Philosophisch-historische Klasse no. 7) (Berlin 1917). PLRE 2.921–22 Procopius of Gaza died aged 62 ca. 528. According to PLRE, he is known to have composed extant works within the period 501–26, which is my guess for the date of this oration, without reading the specialist literature.
Dating summary: range 501–526, midpoint 513.5, class x (historical text), publication 3/3.
S10b Latrines
13ASI Sagalassos (Lower Agora): A substructure of the baths, accessible from the road just off the Lower Agora, was converted into a latrine for ca. 50 visitors, with repairs post-dating the early 6th c. earthquake: Sagalassos Internal Report (2004) 19. This new latrine is thought to have been needed because the ‘earthquake of ca. 500’ put the original latrine of the baths out of use, on which see the summary of F. Martens, Interdisciplinary Research Concerning the Urban Development of Sagalassos. Settlement Development, Urban Layout and Infrastructure (Ph.D. diss., KULeuven) 603. On the earthquake, attested nearby at the Agora Gate, see appendices S6 and F7b. Although I have not seen definitive evidence inside the baths for earthquake damage, ca. AD 500, I will provisionally accept this phasing, as it is attested elsewhere, as part of a poor date. The stratigraphy from within the latrine suggests it was abandoned within the 6th c. It was converted into a dump for manure, from which secondarily dumped layers were recovered, with finds that included unspecified ceramics going into the 6th c.: J. Baeten et al., “Faecal biomarker and archaeobotanical analyses of sediments from a public latrine shed new light on ruralisation in Sagalassos, Turkey”, JAS 39 (2012) (1143–59) 1144.
Dating summary: range 500–600, midpoint 550, class Cs2 (catch all other, earthquake), Cs9 (contextual ceramics), publication 2/3. Poor.
S11 Statues, New or Repaired
13ASI Sagalassos (Upper Agora) (3): A number of Early Imperial statues were re-erected or moved around the Upper Agora within Late Antiquity. This is clear from their distribution, which does not represent an Early Imperial statue landscape: civic honorific statues are concentrated in the rebuilt nymphaeum (although they do not fit the spaces they were found in), whilst statues of earlier emperors are concentrated in one part of the plaza. At the same time, large honorific columns were left in place, as were two statue bases set on pyramids of steps (‘stepped monuments’): L. Lavan, “The agorai of Sagalassos in Late Antiquity: an interpretive study”, in Field Methods and Post-Excavation Techniques in Late Antique Archaeology, edd. L. Lavan and M. Mulryan (Late Antique Archaeology 9) (Leiden-Boston 2013) (289–353) 324–26.
A wall on the north-east side of the Upper Agora includes large numbers of statue bases, which had never before been used as building material in such a high concentration. This suggests that there was an unprecedented tidying up of the lower part of the agora at the time the wall was built. The date of the wall can be put in the period 475–525, based on spolia use, contextual pottery and contextual coins: see appendix F7b. Amidst all of this change, we can only positively prove three cases in which statues were moved to be re-displayed on the plaza surface within Late Antiquity. (i) Firstly, we have a statue base for Ias, a hexagonal Early Imperial base which was re-erected on a non-matching support (made of a seat) in the south-west corner of the agora. (ii) and (iii) Secondly, two blank hexagonal statue bases, set on identical socles, were redisplayed, to stand in a line between two ‘stepped monuments’ (pyramids of steps) on the north-west side of the agora, which themselves supported hexagonal statue bases. The relocated blocks, only seen on the Upper Agora, were likely taken off two adjacent ‘stepped monuments’ when these step pyramids were removed from the plaza.
Of dating, this removal probably occurred after 475, because the wall on the north-east side of the Upper Agora, with large numbers of statue bases, includes parts of exedrae seats. A cross was carefully carved onto the socle of one of the reused hexagonal bases, perhaps to mark its ‘relocation’, which suggests a date after 400, when crosses start to appear on public monuments and inscriptions: see dating foreword. I do not want to evoke the ‘earthquake of AD 500’ for these events, as there is no evidence from the Upper Agora of seismic damage or reinforcement relating to this event. Although there is nothing to date the re-erection of the base of Ias, daughter of Krateros, other than use of reused material, it is highly likely that the statue was re-erected on a seat derived from agora furniture, displaced at the same time as other exedra, which ended up in the wall of 475–525. Thus, the re-erection of the Ias statue likely dates from the same period as the redisplay of the stepped monuments. Find-spots for the bases of Ias and its support, and for the hexagonal bases and their matching supporting blocks, come from excavation notebooks, available to me during my post-doctoral fellowship at Leuven. For a map see Lavan (2013) 321 fig. 10b with 322 table 2. I take the full dating range given for the wall as representing the date for this statue movement.
Of subsequent history, the relocation of the hexagonal statue bases was not the last chapter. Whilst one of the relocated bases was found standing next to its socle between the two surviving ‘stepped monuments’, the second base had been separated into two parts. The distinctive supporting socle had been dragged a few metres east and built into a basin wall on the west edge of the agora, though not before it had been marked with the decorative cross mentioned above, at the very centre of one side, implying that it had been displayed within the Christian period. This basin was established in the period 450–575, according to contextual ceramics of phase 8 Sagalassos Red Slip Ware obtained from the destruction layers beneath an adjacent floor, which contained the pipes which led to this basin (see appendix Y2: phase 3). For full details, see Lavan (2013) 316–23 with figs. 8–10 especially. This very localised spoliation, not entirely completed, suggests that the later 6th c. occupation of the agora did not respect its monuments in the same way that the earlier 6th c. occupation did, even if much of its classical monumentality survived until the abandonment of the city in the first decades of the 7th c.
Dating summary: range 475–525, midpoint 500, class Cs5 (catch-all specific spolia), Cs9 (contextual ceramics), publication 3/3.
See also (appendix D2): 16ORI Caesarea Maritima, on the esplanade created sometime 546–614 by the governor Fl. Strategius, flanking the wall-arch to the south, were two colossal reused statues of two emperors, one in porphyry (sitting on a granite throne), which was broken when displayed, indicated by the “makeshift bases and odd bits of stone used to prop the statues up where parts were missing”,
See also (appendix K1a) 12CPL Constantinople (Forum of Constantine): A gilded equestrian statue (on columns?) was erected to honour general Nicetas, cousin of Heraclius (LSA 478) (dated to AD 614–17).
See also (appendix K7a): 13ASI Aphrodisias (South Agora): A fragmentary marble plaque, from the late basin of the Agora Gate, honoured Flavius [Justinia]nus, given by an unknown dedicant (ALA 81 = LSA 231). Roueché believes the letter style is 6th c., based especially on the forms of three letters. She used the name of the emperor Justinian (reigned 527–65) to fill the gaps, as this name fits the space for the number of letters available. This base was perhaps displayed on the agora, though not necessarily on the fountain, into which it might have been thrown.
See also (appendix S2): 09DAC Justiniana Prima: A stone support for a column stood at the centre of the plaza, which may have held the bronze statue (for an emperor given the military dress) of which two fragments were found nearby; 10MAC Dyrrachium: At the centre of the plaza a round foundation, 5.75 m in diameter, supported a square stone plinth of 1.82 m by 1.8 m, composed of two massive stone blocks, which might have supported a statue of Anastasius, as this city was his birthplace, perhaps set on an honorific column. The plaza is dated to 500–525.
See also (appendix S3): 12CPL Constantinople (Strategion): After a fire in the Strategion, recorded by Marcellinus Comes in AD 510, the arm of [bronze] statue holding a horn of plenty, set above an arch was immediately restored (so we can give a date of 510).
See also (appendix S4): 13ASI Aphrodisias (South Agora): A base for an honorific statue was erected to clarissimus Albinus, native of Aphrodisias and evergete, by the city; it was found in the west portico, at the southern end (ALA 82 = LSA 232). It is dated to the 6th c. in ALA but recent discussion of associated acclamations to this person, inscribed on the columns of the same west portico, suggests a date of 527–614 (i.e. from time of Justinian to Persian invasion of Asia Minor).
See also (appendix F9): for statues on honorific columns: 06ITS Rome (for Phocas) dedicated in 608; 12CPL Constantinople (5) (for Anastasius in 505–506, Justinian in 543–44, Justin II in 578 (not completed), Theodora in 527–65 and Phocas in 609); from 15ORI Scythopolis? in 500–535; 16AEG Abu Mina? probably within the period 475–532. I do not think these are comparable to ordinary statue monuments, but include them here because there are so few statues in fora / agorai at this time.
S12a Civil Basilica Repairs
See also (appendix K4a): 15ORI Antioch in Syria (Hellenistic Agora): ‘Basilica of Rufinus’ (likely a double-aisled portico) rebuilt in 507–18.
See also (appendix X1a): 07AFR Carthage: Later building, after 5th c. decay, included the provision of two massive new apses at either end of the basilica and new capitals, which are apparently in a reconquest period style. Although the provision of apses at each end of the building would be consistent with its continued use as a civil basilica, a change of function has been suggested, as a fortified convent was built after the Justinianic reconquest by the general Solomon near to the Mandracium harbour. Dating is of 533–58.
See also (appendix X1b): 13ASI Ephesus: The Hall of Nero, on the Lower (tetragonal) Agora, saw a number of building interventions, of which a late phase in the same ‘tribunal’ area saw the installation of a peristyle court in the mosaic room. This has been associated with the second general phase of renewal of the agora, dated between ca. 550 and 610. There is no direct structural argument that can at present link the tribunal area to these changes, except that the tribunal area of the Hall of Nero and north portico of the agora are adjacent. However, I have noted in appendix K2b that the distribution of spolia in the agora suggests that the front of the east portico shops, which would have also served as the west wall of the Hall of Nero above, was probably rebuilt as part of the same later 6th c. works. Judicial inscriptions carved on the walls of the basilica (the earliest dated text being of 569) suggest that law courts were being held here.
S12b Curia / Bouleuteria and Other Offices, Repairs, 6th-7th C.
10MAC Philippi: The façade of the curia was given wooden props at some stage, as revealed by a series of holes in the paving of the façade area and two ‘massifs’ (likely masonry buttresses), which were added to the south wall: M. Sève and P. Weber, Guide du forum de Philippes (Sites et monments 18) (Athens 2012) 66–67. I noticed that the curia has obvious reused material, though not architectural spolia, in its south wall on the south side and also I think in the north wall, where the masonry is mainly rubble and a few bits of brick, poorly coursed. This is however different to other late masonry on site, notably that which is associated with phase 3 of the forum, which tends to be rougher: L. Lavan site observations, April 2017.
Of phasing and dating, the first set of features (not those noted by me) are ascribed to phase 3 of the agora, thought to represent a response to an earthquake [not identified, though see candidates in Guidoboni Catalogo 688–704]. This is reasonable, given that the buttresses are reinforcements to a weakened structure, and reinforcements are known to the adjacent basilica, with walls being built between columns: Sève and Weber (2012) 65 (basilica details). However, this phase of rebuilding is dated by me about a half-century later than Sève and Weber, to the second half of the 6th c., from the impost capitals used in the rebuilt south colonnade of the forum, which can be paralleled with dated examples from Aliki, and from the association of the structure with work on two churches, inspired by Constantinopolitan great churches finished in 536–537 (SS Sergius and Bacchus; St. Sophia) at either end. A terminus is the regional downturn that accompanies the Slav invasions, which were decisive by 616: see appendix S4 with appendices C4 and E1 for the dating of the third phase of the forum. The attribution of the curia repairs to the third phase of the plaza must remain suggestive, in the absence of independent dating evidence.
Of the earlier repair / construction work I identified, one might envisage a missed late 3rd to 5th c. phase of building work in the curia, but I am not confident in their dating. Such very slight reuse might have taken place in the 2nd or early 3rd c.
Dating summary (repairs to curia in guidebook): range 536–616, midpoint 576, class Cs3 (associative, phase of development), Cs1 (architectural style), z (regional development), publication 3/3. Poor.
12CPL Constantinople (Senate in Augusteion): Justinian rebuilt the Senate in the Augusteion. The new building is described by Procopius: Procop. Aed. 1.10.6–9. He notes that it stood on the east side of the agora where Justinian had his column. The building is sometimes identified by scholars with the Magnaura, an audience hall located nearby, although a passage of Chron. Pasch. Olympiad 327 AD 531 (dating from edn. of Whitby and Whitby (1989) although clearly Januaury 532 from Theophanes A.M. 6024 AD 531/32 edn. of Mango and Scott (1997)) refers to the Magnaura as well as to the Senate, within the same riot, suggesting they were different. On this difficult debate see J. Kostenec, “Palace of Magnaura”, in Encyclopaedia of the Hellenic World, Constantinople (2008), http://www.ehw.gr/l.aspx?id=12436 (last accessed March 2016).
Of sources, we have both the extensive description of Procopius and antiquarian drawings, which show a row of columns on the south side of Hagia Sophia, apparently aligned north-south. Of antiquarian evidence, Petrus Gyllius also saw 7 Corinthian columns near to the pedestal of the column of Justinian, for which he gives measurements: Petri Gyllii, De topographia Constantinopoleos, et de illius antiquitatibus libri quatuor (Lyon 1562) 2.18 p. 111 = J. Ball transl., Petrus Gyllius (Pierre Gilles), The Antiquities of Constantinople (London 1729) 2.18 p. 137. These columns, 6 in number, were also seen next to the column of Justinian by C. Buondelmonti, Liber Insularum Archipelagi (no place 1420) 65 (G.R. Ludwig von Sinner ed., Christoph. Bondelmontii, Florentini, Librum Insularum Archipelagi (Leipzig and Berlin 1824) 122). Given that Procopius records a porch of 6 columns, with 4 projecting out, it is fair to think that the columns seen by Gyllius and others represent the Senate porch.
Of dimensions, the depth of the building likely measured ca. 63.667 m or 70.334 m long, from front porch to back wall, or 57 m if the porch sat within the Augusteion’s east portico line, rather than projecting from it. These ridiculously ‘accurate’ lengths are speculative and result from combining figures from Gyllius with an estimate of 57 m derived from my calculations, which assume a square plaza set around the column of Justinian (see appendix S3). The width of the porch seems likely to be ca. 24.688 m, based on the spacing of 4 columns being the same, which are reported by Gyllius as standing about 20 ft and 10 digits from each other [20 by 0.299 m + 10 by 0.0187 m = 6.167 m] (see below). This measurement assumes that Gyllius was measuring from the centre of each column. Gyllius’ accuracy (if not infallibility) can be confirmed from his detailed measurements of the Column of Arcadius. See appendix F9 for conversion of his non-imperial measures into metres.
Of architectural form, we can note Procopius’ description in the form of a prostyle porch with an internally-arcuated pediment (likely a ‘Syrian pediment’):
‘To the East of this marketplace stands the Senate House, surpassing description by reason of its costliness and every element of its construction, the work of the Emperor Justinian … Six of its columns stand in front of it, two of which have between them the wall of the Senate House which faces the West, while the four others stand a little beyond it … And the columns form a porch (stoa) which carries a roof curving into a vault (tholos), and the whole upper portion of the colonnade is adorned with marbles which rival the columns in their beauty …’
We can suggest a porch height based on using the colonnade height from Gyllius (below) of 13.9 m compared with the Syrian pediment of the porch of the Theodosian Hagia Sophia from the reconstructed section of A.M. Schneider, Die Grabung im Westhof der Sophienkirche zu Istanbul (Istanbuler Forschungen 12) (Berlin 1941) (tafel 4) [which is based on the study of surviving fragments and a surviving ground plan]. Roughly hand-measuring off this drawing, the Hagia Sophia porch height of the vault above the floor seems to be ca. 9 m and the height of the pediment above the floor seems to be ca. 10.95 m, for a colonnade height of 6.95 m. Transposed to the Senate House colonnade height, with the same ratios, the Senate House vault height above the floor is ca. 18 m and the pediment top height above the floor is ca. 21.91 m.
Of architectural features, Gyllius reports Corinthian columns, which presumably means they were still topped by Corinthian capitals. Procopius says that the columns are “white in colour, and in size, I believe, they are the largest of all columns in the whole world”. Gyllius says that each of the columns were 30 foot and 6 digits high [30 by 0.299 m + 6 by 0.0187 m = 9.0822 m], with a total colonnade height for column, capital and pedestal of about 46.5 foot [46.5 by 0.299 m = 13.9035 m], standing about 20 foot and 10 digits from each other [20 by 0.299 m + 10 by 0.0187 m = 6.167 m]. This massive colonnade, which is shown as 6 columns on some early drawings of the city, is thought by Stichel (2000) to represent the façade of the Justinianic Senate House (see below). In the latter arrangement, the columns would represent a projecting building façade, and so need not represent the standard portico height of the rest of the plaza, although they could still be integrated within the line of the portico, as was the ‘Syrian pedimental’ porch of the Theodosian Hagia Sophia. In the latter case, the Syrian pediment’s 4 front columns are not equal in their spacing, as they allow a wider central passage, but this is not the case at Diocletian’s palace of Split, where they appear broadly equally spaced. Gyllius reports the presence of an inscription on one of the columns with the name of Constantine, a cross and ‘by this sign conquer’ in Greek.
Of statuary, Procopius remarks that “the roof is wonderfully set off by a great number of statues which stand upon it.”
Of phasing, I must confess to a lingering doubt that the Syrian pediment of the porch and the statues on its roof seem likely to be a survival of the Theodosian Senate House rebuilt in 414 (see appendix L2), rather than anything built in the time of Justinian. Whilst wall arches featuring arcades are known in the period, projecting pedimental porches of this style are more frequent in the 4th and 5th centuries. The popularity of arcuated lintel pediments does extend further into the period, notably being used for the Theodosian Hagia Sophia, of 404–15, which is the latest I know of in real buildings (see appendix C4). Its use in art continues longer, appearing (without the top of the pediment, just the arcuated lintel) in the David Plates, thought to celebrate the victories of Heraclius over Persia (so likely 629–34, after the defeat of Persia, but before the Islamic invasion of Syria): M. Parada López de Corselas, “The arcuated lintel and the ‘Serlian Motif’. Imperial identity, architectural and symbolic interactions in ancient Rome”, in SOMA 2012. Identity and Connectivity: Proceedings of the 16th Symposium on Mediterranean Archaeology, Florence, Italy, 1–3 March 2012, vol. 1 (Oxford 2013) (479–86) 482. The statues on the roof also seem more appropriate to the former period, even if imperial statues continue into the reign of Justinian and Justin II at Constantinople. Thus, it is possible that the Justinianic rebuilding kept part of the shell of the Theodosian Senate House, which had been rebuilt in a style comparable to the porch of the Theodosian Hagia Sophia, although on a much larger scale. However, in this study I concentrate only on dating the rebuilding activity of Justinian, and leave this as a doubt.
Of dating, we can reasonably assume that the Senate House, restored by Justinian, was finished in time for the construction of the emperor’s column that faced it, the set piece of Procopius’ description: Aed. 1.10.5–9. This is likely to have been erected as the crowning event of the construction. Theophanes gives AM 6036 (AD 543/44 date in edn. of Mango and Scott (1997)), 7th indiction, the 17th year of Justinian, for the erection of the column, a regnal year also given by Zonaras. It is given as the 16th year and 7th indiction by Cedrenos (edn. of Bekker (1838), vol 1. p. 656). A safer TAQ for the works is 558. Procopius’ text (in which the plaza seems complete) likely dates to before the collapse of Hagia Sophia’s dome in 558, as acknowledgement of this does not feature in it, despite a description given to the monument in 1.1.66–78, which discusses problems of the stability of the dome, which would be odd, given the expectations of the reader after 558. However, the speed of building at Hagia Sophia, of only 5 years from destruction to inauguration, suggests that 543/44 is a reasonable TAQ. It is preferable to use the date from the column erection rather than the fire rebuilding speed, as this is a specific piece of info related to the same complex.
Dating summary (for Senate rebuilding with restoration of the plaza by Justinian): range 532–44, midpoint 538, class x (historical texts), publication 3/3.
See also (appendix S12c) 13ASI Aphrodisias: Some of the repairs listed immediately below in the period 460–614 could relate to the traditional use of the bouleuterion, as the reuse for entertainments does not rule out continued political meetings, even if there are no political acclamations or 6th c. statue dedications that might suggest this.
S12c Curiae / Bouleuteria etc., Disuse, 6th-7th C.
13ASI Aphrodisias: The bouleuterion-odeon was modified into a facility for entertainments. The last three rows of seats were replaced with a brick podium, which now separated the audience from the pit of the orchestra, in the manner of theatre-amphitheatre conversions. The new podium wall was given the same revetment as the proscaenium. Above the new podium, a walkway was inserted, composed of reused seats and other blocks, of the same type that were used to plug a hole in the opus sectile floor of the orchestra. In a gap between this sectile floor and the new podium was a row of low rough blocks, which may have supported temporary seating of some other feature, which extended over the floor. This can be deduced, because steps (made of reused seats), now established to link the new podium to the orchestra, ‘floated’ a little above the floor, as if some features were missing. The cornice of the proscaenium was also reset, made up of moulded blocks of different origins: L. Bier, “The Bouleuterion”, in Aphrodisias Papers 4. New Research on the City and its Monuments, edd. C. Ratté and R. R. R. Smith (JRA Supplementary Series 70) (Portsmouth, R.I. 2008) (144–168) 163–64.
Other late features, spatially separate from the orchestra, but perhaps of the same phase, are 4 new staircases, replacing two earlier ones, now giving access to the upper cavea: Ratté and Smith (2008) 165. I observed that at least one of these staircases was made with reused materials: L. Lavan site observation, April 2001. There were also short staircases made inside the stage building including reused material. They were covered with plaster, painted with orange wavy lines to imitate marble: Bier (2008) 165. Indeed, Bier (2008) 159 suggests that the stage building might have been rebuilt within Late Antiquity, although he provides no evidence apart from mismatching statuary. Finally, there are two different systems of holes arranged within the cavea seating. These holes seem to have supported some kind of awning, holding either posts or ropes. They may have dated from a time when the roof was no longer present, so thus being a late detail: Bier (2008) 166. The lack of drains in the cavea has led to the suggestion that it might have served as a reflection pool, although there is no positive evidence to support this: Bier (2008) 116.
An inscription from the new parapet of the stage building refers to building work on a palaestra (τὸ ἔργον τῆς παλαίστρας) by an unnamed scholasticus and pater, which can be taken as referring to the building works described above: C.M. Roueché, Performers and Partisans at Aphrodisias (JRS Mon. 6) (London 1993) no. 47 = ALA 43. The office of pater puts the date of the works after ca. AD 460, when this office is mentioned in relation to public building: C.M. Roueché, “A new inscription from Aphrodisias and the title πατὴρ τῆς πόλεως”, GRBS 20 (1979) 173–85, esp. 182. However, the works should date from before the decisive Persian invasion of Asia Minor in 614, after which civic secular building work is not known in the region. Further inscriptions, found on the cavea seats, suggest that the building was used for either assemblies or entertainments: place inscriptions were recorded for ‘the younger men’, ‘the Blues’, ‘the Jews’ and the ‘old Jews’, which does not look like anything to do with the city council, and in the case of the ‘Blues’ suggest a date in the 5th or 6th c. AD: Roueché (1993) 117f. and 123.
Dating summary (late building works): range 460–614, midpoint 537, class Cs6 (absolute, inscription in situ), z (regional development), publication 3/3.
15ORI Pella in Jordan: A great staircase serving the ‘phase C church’ contains reused seats from the adjacent odeon. This was a building that faced onto the presumed agora of the city and so may have had the function of bouleuterion. The reuse of the seats suggests that the odeon was ruinous at the time. The ‘phase C church’ however cannot be dated (I disregard local stylistic arguments), except to place it before a site-wide earthquake of 749 [Guidoboni Catalogo no. 220] and after a coin of Arcadius dated to 395–408 from the phase A church. This gives us a date range of 395–749 for the church: see appendix Y5. Inside the ruined odeon, stripped of much of its decoration, was recovered a dump from the late 6th and 7th c., which included a “significant corpus of Late Byzantine potsherds”: R.H. Smith and L.P. Day, Pella of the Decapolis 2. Final Report on the College of Wooster Excavations in Area IX, the Civic Complex, 1979–1985 (Wooster, Ohio 1989) 8. I was not able to access the pottery report, but this suggests a provisional contextual TAQ for the disuse of the odeon, of ca. 600, using a deposition start date of the late 6th c. onwards. I am assuming this is a primary rubbish deposit, accumulating gradually in this spot, rather than a mixed single load containing late 6th and 7th c. pottery dumped here.
Dating summary (disuse of odeon): range 395–600, midpoint 497.5, class Cs9 (contextual ceramics), publication 2/3.
See also (appendix U2): 06ITS Rome: The Senate Building was converted into a church during the period AD 625–38.
See also (appendix X2): 15ORI Scythopolis: The Odeon-Bouleuterion was demolished to build the Sigma Plaza in AD 506–507.
T1a Market Buildings
13ASI Sagalassos (Upper Agora, Market Building): A reorganisation of the rooms divides the earlier central space into smaller rooms (5–7). This development is given a contextual date by the ceramic assemblage from a foundation trench (room 7, layer 7). These ceramics are of Sagalassos Red Slip Ware phase 8 (450/75–550/75 AD). The foundation trench did not contain phase 9 (550/75–ca. 700 in 2013) ceramics, unlike the subsequent fill, which raised the floors of these rooms: Internal Report (2004) 8–14. Given the lack of detail about specific types of pottery present within these classes, I accept the full date range of phase 8 pottery as providing an approximate contextual date for this reorganisation.
Dating summary: range 450–575, midpoint 512.5, class Cs9 (contextual ceramics), publication 3/3.
See also (appendix W1): 13ASI Sagalassos (Upper Agora, Macellum): The rebuilding of shops on all three sides of the macellum currently dates between 487.5 and 575, based on my analysis of the report, although recent reports talk of the rebuilding work as being 5th c.
T1b Shop Functions on Fora / Agorai
13ASI Ephesus (Lower Agora): Amongst the evidence for a glass workshop, of apparent 6th c. date, were glass weights with monograms; only preliminary reports are published so far: F. Krinzinger, “Ephesos 2000”, Kazı Sonuçları Toplantısı 23.2 (2001) (127–35) 129. The workshop has produced coins of the time of Justinian, but not apparently later, being set at the higher [later] 6th c. level of the agora [550–610], for which see appendix S3, plus appendices S4, S6 and S10a: S. Karwiese, “Agora: ein Münzteppich”, in: S. Karwiese et al., “Ephesos”, ÖJh 67 (1998) Grabungen 1997, 11–12. I give here a date of 550–600 for occupation, as the report does not anticipate any early 7th c finds.
Dating summary: range 550–600, midpoint 575, class Cs8 (contextual coins), Cs3 (associative, phase of development), publication 3/3.
See also (appendix Y2): 13ASI Sagalassos (Upper Agora, west side): Shops on the west side of the agora produced important evidence of their final occupation level for room 1 (of 525–641) and for room 16 (525–75). In room 1 was found a set of professionally-inscribed metal weights for checking the value of currency and weights, possibly indicating some administrative function, like a market official or perhaps a tax office?. In room 16 were found smith ‘hearth cakes’ (of vitrified waste from a blacksmithing hearth bottom), metal wire, charcoal and hammerscale fragments, indicating the presence of a blacksmith’s and metal workshops. 13ASI Sagalassos (Lower Agora, west side) (2): Within a row of shops occupied within the period ca. 500–610, in W6 and W7, was found a ceramic assemblage, indicating the cooking and serving of food, alongside a large concentration of coins. These finds suggested the selling of food to passers-by, therefore a restaurant function. The similarity of the ceramic assemblage of W2/W3 suggested a similar function here too, as did a fireplace in W3.
T2 Temples, New and Repaired
12CPL Constantinople (Tychaion on ‘Basilica’ courtyard): Al. Cameron has suggested that the τετραπόροις ἁψῖσι (translated as ‘4 porticoes’ previously, but now as ‘4 arches’), built by Theodorus, thrice-prefect on the portico of the Basilica, could have been a cross-vaulted 4-column structure, which might have acted as an entrance arch: Anth. Graec. 9.696 with 9.697. Cameron identified this Theodorus with an urban prefect under Justin I, who was in his third prefecture in AD 520. This same prefect (9.697) also decorated the Temple of Tyche in the same ἁψίς (arch), perhaps providing columns, to which the epigram refers adjectively. For discussion, see Al. Cameron, “Theodorus τπισέπαπχος”, GRBS 17 (1976) (269–80) 269–73; PLRE 2.1096 Theodorus qui et Teganistes 57. The linguistic nuances need not imply a monumental arch. Rather, it should be remembered that some eastern tychaia were canopy monuments supported by columns, see e.g. P. Talloen, Cult in Pisidia. Religious Practice in Southwestern Asia Minor from the Hellenistic to the Early Byzantine Period (Ph.D. diss., Univ. of Leuven 2003) Text vol. 1. 69–70, vol. 2. 103, Catalogue vol. 2 no. 573, fig. 220. So, it is possible that the inscriptions both relate to the same monument: a canopy / ciborium-type tychaion restored in 520. See discussion of the tychaion of Constantinople with sources in F.A. Bauer, Stadt, Platz und Denkmal in der Spätantike. Untersuchungen zur Ausstattung des Öffentlichen Raums in den Spätantiken Städten Rom, Konstantinopel und Ephesos (Mainz 1996) 219 and G. Dagron, Constantinople. Naissance d’une capitale (Paris 1974) 373–74.
Dating summary: 520, class x (historical text), publication 3/3.
12CPL Constantinople (Augusteion): The same Theodorus, thrice-prefect, also built a temple / shrine (νηός), honouring Justin I and also Justinian. The latter was honoured as a general and as a son of Justin I, so before Justinian became emperor. The construction is known from epigrams (recorded on inscriptions) in Anth. Pal. 1.97–98. These texts are placed ἐν τῇ Μελέτῃ by a lemma of the Anth. Pal., a place which seems to be identical to the Gate of Melete, mentioned in Cer. 1.2 (as being near the place called Achilles) and Cer. 1.8 (as between the Chalke and the Milion, when it is called the ‘Great Gate of Meletè’ (πύλη τῆς Μελέτης)). See C. Mango, The Brazen House: a Study of the Vestibule of the Imperial Palace of Constantinople (Copenhagen 1959) 73–78, and p. 23 (for a reconstructed plan of the area). Given the position of the gate in this street, surrounded on one side by the Great Palace / Zeuxippos and on the other by the Augusteion, it is very likely that the Gate of Melete was a gateway into the Augusteion plaza. The shrine is described by the Anth. Pal. as having been built by the consul Theodorus, three times a prefect, which is a man of this name who was prefect for the third time in 520: see Al. Cameron, “Theodorus τρισέπαρχος”, GRBS 17 (1976) 269–80; PLRE 2.1096 Theodorus qui et Teganistes 57. The structure is described by Anth. Pal. 1.98 as ‘glittering with a lustre of vast stores of minerals’, which suggests a mosaic. It is difficult to know what kind of structure is implied, but it sounds like a sanctuary for imperial statues, perhaps similar in size to the canopy tychaion described above, with scope for a mosaic in its vault. However, no statuary is referred to. One might alternatively suggest that the sense of νηòς was not a temple, but a ship monument (as the language would allow) that was added to the Liburna marmorea, navalis victoriae monumentum (Not. Const. Region IV). This monument may have been in the Augusteion, as, in the Nika Riot of 532, the rioters coming from the Octagon (of the ‘Basilica’ courtyard) set fire to the Liburnon near the Magnaura (perhaps an audience hall on the north-west side of the palace), on which see: Chron. Pasch. Olympiad 327 AD 531 (dating from edn. of Whitby and Whitby (1989), although clearly January 532 from Theophanes A.M. 6024 AD 531/32 dating in edn. of Mango and Scott (1997)), plus (Mango (1959) 56–57 vs. J. Kostenec, “Palace of Magnaura”, in Encyclopaedia of the Hellenic World, Constantinople (2008), http://www.ehw.gr/l.aspx?id=12436 (last accessed March 2016).
Dating summary: 520, class x (historical text), publication 3/3.
T3 Temples, Demolished
06ITS Rome (Forum of Augustus): Patricius Decius demolished the temple of Mars Ultor in the Forum of Augustus, probably in the later 5th c. An inscription ‘pat(rici) deci’ testifies to this on a drum of one of the columns found adjacent to the temple that seems to have come from the peristasis: R. Meneghini and R. Santangeli Valenzani, “Episodi di trasformazione del paesaggio urbano nella Roma altomedievale attraverso l’analisi di due contesti: un isolato in piazza dei Cinquecento e l’area dei fori imperiali”, Archeologia Medievale 23 (1996) (53–99) 78–81 with p. 80 fig. 21. The dating of the inscription to Late Antiquity depends on the form of the ‘A’, but also because a family of Decii is known to have been prominent in Rome at this time, of which two individuals are known to have used the title patrician with Decius as their principal name. These are PLRE 2.349 Caecina Mavortius Basilius Decius 2 (urban prefect, praetorian prefect, and consul in 486) and PLRE 3.391 Decius 1 (consul in 529, who fled Rome for Constantinople in 546, when the armies of Totila captured the city). The inscription is in the genitive, likely indicating that the column belonged to Decius. Given the prominence of these two individuals, the authors of the report suggest a date in the late 5th c. or first decades of the 6th c., the years of Gothic rule. The career order of Decius 2, as reconstructed in PLRE, from CIL 10.6850/1 = ILS 827, suggests that he was made patrician after his consulship in 486. Thus, we can contemplate a range for the demolition from 486–546.
Dating summary (for demolition): range 486–546, midpoint 516, class Cs6 (absolute, inscription almost in situ), publication 3/3.
07AFR Lepcis Magna (Old Forum) (2): Two temples (of Liber Pater, and Rome and Augustus) were covered by the line of a 6th c. sea wall, in such a way as it ran through their cellae and over the partly demolished podia: D. Pringle, The Defence of Byzantine Africa from Justinian to the Arab Conquest, part 1 (BAR-IS 99) (Oxford 1981) 209; R.G. Goodchild and J.B. Ward-Perkins, “The Roman and Byzantine defences of Lepcis Magna”, BSR 21 (New Series 8) (1953) (47–53) 54 (general map), 56, 59. See R.B. Bandinelli, E.V. Caffarelli, G. Caputo, and F. Clerici, The Buried City: Excavations at Leptis Magna (New York 1966) 84 fig. 235 for a map of the line of the wall in the forum.
Of phasing, there are various theories on the collapse or earthquake damage to the temples, which do not seem to rest on archaeological dating evidence: see works listed and discussed in M. Pentricci, “L’attività edilizia a Lepcis Magna tra l’età tetrarchica e il V secolo: una messa a punto”, in Leptis Magna. Una città e le sue inscrizioni in epoca tardoromana, edd. I. Tantillo and F. Bigi (Cassino 2010) (97–171) 141. However, Pentricci makes the observation that the wall rests on a thick level of masonry that had to alter its level as it passed the walls of the temples, suggesting that they had already collapsed [I did not quite understand the observation made here]. Thus, we might consider a significant gap between the temples collapsing and the fortification wall being built. Although I can accept that the temples might have been degraded with the removal of decoration etc. before they were covered by the wall, I have not yet seen any evidence to suggest there was an earlier intentional demolition prior to the wall being built, on a forum where one other temple survived to be converted into a church.
Of dating, Pringle (1981) 211 considers that the initial fortification of the city was complete by 543, as this city seems to have been fortified in Procopius’ account when the leaders of the Leuthae were slain at a banquet held by the dux Tripolitianae Sergius: Procop. Vand. 4.21.2–15. The initial fortification should thus probably date after the designation of the city as seat of the Dux in 534: Cod. Iust. 1.27.2.1a (AD 534). However, the fortification that runs through the old forum is a later addition to the circuit, marking a reduction in the defended area. Yet, there is no reported difference in its construction technique, apart from the fact that it is less straight than the original fortification, according to Goodchild and Ward-Perkins (1953) 59. Thus, the inner fortification can be given the same date as the outer one, based on phase of development, because other changes in the circuit are reported, which show a strong similarity in construction technique: Goodchild and Ward-Perkins (1953) 72. One might like to credit this fortification redesign to after the events of 543, as this might provide a context for a replanning, but there is no argument under the rules of this study that can do this. Frankly, we do not have a rounded picture of the challenges facing the defenders of Lepcis in this period.
Overall, I stick to a date range for the fortification of 534–43, and by extension as the TAQ demolition date for the temples, not being certain that there was any active demolition before this.
Dating summary: range 534–43, midpoint 538.5, class x (historical texts), Cs3 (associative, phase of development), publication 3/3.
T4 Temples, Converted into Other Uses
13ASI Side (Harbour Square): The double-temple complex on the Harbour Square (probably the real agora of the city) was converted into a church during Late Antiquity. The temples themselves were not used as part of the church buildings directly, but incorporated within the atrium, whilst a propylon of the enclosure was enclosed within a basilical church. See P. Talloen and L. Vercauteren, “The fate of temples in late antique Anatolia”, in The Archaeology of Late Antique Paganism, edd. L. Lavan and M. Mulryan (Late Antique Archaeology 7) (Leiden 2011) (347–87) 364–65 with fig. 5 (plan), drawing on A.M. Mansel, Side. 1947–1966 Yılları Kazıları ve Araştırmalarının Sonuçları (Türk Tarih Kurumu Yayınları V.33) (Ankara 1978) 121–35 and 257–66 with plans on p. 122 fig. 133 and p. 258 fig. 284. The columns of one of the temples were retained to serve as part of the colonnade, or perhaps a propylon, whilst the other parts were demolished: C. Foss, “The cities of Pamphylia in the Byzantine age”, in Cities, Fortresses and Villages of Byzantine Asia Minor (Collected Studies Series 538) (Aldershot 1996) 39, with R. Bayliss, Provincial Cilicia and the Archaeology of Temple Conversion (BAR International Series 1281) (Oxford 2004) 40 for the suggestion that one temple served as a propylon.
Of dating, the only credible suggestions come from descriptions of the capitals (despite claims of the church being 5th c. by Mansel or earlier by others). Impost capitals, found recently, suggest a dating to the first decades of the 6th c. for G. Grassi, “Scultura architettonica e spolia marmoree della Panaghia di Antalya nel quadro della produzione artistica dell’Asia Minore meridionale in epoca paleobizantina”, Milion 2 (1990) (73–134) 95 with plate 38 figs. 46–47. Grassi describes the imposts as being decorated with a simple Latin cross, whilst “un ovolo inquadrato da fogliette si staglia tra le profonde volute sottanstanti”; only one of the imposts is richer, as the cross is surrounded by “girali” (which I take as meaning ‘plant scrolls’, though perhaps it should be ‘tendrils’). On the photograph, the capitals are labelled as Ionic impost capitals, which seem deeply chiselled, showing two volutes and a central ovule, the latter being surrounded by flowers.
The capitals are similar to a number of Constantinopolitan capitals catalogued by Zollt (nos. 49–62 in the plates 14–16), although none are identical. Plate 16 no. 61 is only one in Zollt’s book that is provenanced to a dated church: St. John the Forerunner church in Hebdomon, built ‘a few years before AD 555’. However, the St. John capital almost lacks an ovule, which is only marked out as an ovular gap in the leaves that on other capitals frame the ovule. The capitals in Zollt’s catalogue are all given dates in the 6th c., although with lesser degrees of reliability, which I do not want to go into here. Given that our capitals are slightly different to all of the others illustrated by Zollt, a ‘poor’ date in the 6th c. for the Side church would seem to be justified. The date of St. John’s Church in Hebdomon is based on Procop. de aed. 1.8.15, who describes Justinian’s rebuilding of it as ‘recent’. De aedificiis probably dates sometime before the collapse of the dome of Hagia Sophia in 558, as acknowledgement of the collapse does not feature in Procopius’ text. This is despite a description given to the monument at 1.1.66–78, which discusses problems of the dome’s stability, which would be odd after 558, given the expectations of the reader. B. Croke, “Procopius’ ‘Secret history’: rethinking the date”, GRBS 45 (2005) (405–431) 427–29 has recently argued for a date of composition of de. aed. of AD 554, if not entirely clearly, noting that various historical events alluded to should date the text to after 540 but before 559. He cites previous bibliography.
Dating summary: range 500–600, midpoint 550, class Cs1 (architectural style), publication 3/3. Poor.
See also (appendix S5a): 02HIS Ilici: Church built inside forum temple, sometime in 575–725.
U1 Churches on, but Not over, Fora / Agorai, 4th–5th C.: the West
In this appendix, and in those that follow, I am not concerned with the architectural form of the church, only with its date, as all I care about here is the fate of the fora or agorai. I am normally concerned with the installation of the first church on a forum / agora, not subsequent churches built over its site, except when a second church continues to respect a plaza and thus indicates that it continued in use as an open space.
05ITA Aosta (Augusta Praetoria Salassorum): The church at Aosta was apparently built next to the forum, but I have never been able to obtain an excavation report on the subject: Ch. Bonnet and R. Perinetti, Aosta. I Primi monumenti Cristiani (Aosta 1986) (not seen).
Dating summary: Pending.
07AFR Iol Caesarea (Cherchel): A possible church was built at the side of the forum during the 5th c., out of reused building materials. Its identification as a church is based on the tripartite division of the space leading to an apse, the presence of a raised chancel and on internal fittings. The work is given a TPQ from a latest coin of 410–21, derived from construction levels outside the church, and from mid- to late 5th c. pottery in make-up layers for the same building. Ceramics from the construction levels were of mainly late 4th to mid-5th c. date, with Hayes 67, given as 360–470 [ca. 360–470 in LRP], and 88/81 given as late 4th-mid-5th c. [I have not found this variant, only that 81 is 440–500 in LRP; 88 is 525–612.5 Bon. There might be a mistake here]. The sherds in the make-up levels are as follows: Hayes 91, given as late 4th to late 5th c. [350–450 in LRP/SLRP]; Hayes 76, given as ca. 425–75 [400–75 in LRP/SLRP]; Hayes 88/81 [see above]. No 6th c. pottery was found, except in layers above the church: T.W. Potter, Towns in Late Antiquity: Iol Caesarea and its Context (Oxford 1995) 40, 44, 47, 48. This report is a [fulsome] summary of N. Benseddick and T.W. Potter edd., Fouilles du forum de Cherchel 1977–1981 (Algiers 1993) 2 vols, a publication that is very difficult to obtain. The evidence suggests a contextual date of 25 years after 425 for the construction of the church, based on the start date of the latest find, which is Hayes 76. For a critique of the interpretation of the building as a church: N. Duval, “Une petite église chrétienne nord-africaine sur le forum de Cherchel”, Revue des Études Augustiniennes 34 (1988) (247–66) 249–53.
Dating summary: range 425–50, midpoint 437.5, class Cs9 (contextual ceramics), Cs7 (TPQ coin), publication 3/3.
See also (appendix U2): 07AFR Sabratha: A church was installed in the civil basilica on the forum, after the secular reconstruction of the civil basilica, within the period 400–500.
See also (appendix V1): 04VIE Aix en Provence: The building of a baptistery in the forum in the period 500–525 implies that a church was established in the civil basilica at this time or earlier. This is because the baptistery was raised up on fill to the same level as the building to the north (likely a civil basilica), where the cathedral stands today.
Discounted: 04VIE Arles (Arelate): There is no agreement whether the Basilica sancti Stephani, of mid-5th c. date, was located on the forum. For a review of the discussion see M. Heijmans, Arles durant l’Antiquité tardive de la duplex Arelas à l’urbs genesii (CEFR 324) (Rome 2004) 261–62.
U2 Churches on but Not over Fora / Agorai, 6th–7th C.: the West
06ITS Rome (Forum Romanum): SS Cosma e Damiano was installed inside the so-called audience hall of the city prefect in AD 526–30 (Lib. Pont. 1.279).
Dating summary: range 526–30, midpoint 528, class x (historical text), publication 3/3.
06ITS Rome (Forum Romanum): S. Adriano in foro was installed inside the Senate House in AD 625–38 (Lib. Pont. 1.323–24).
Dating summary: range 625–38, midpoint 631.5, class x (historical text), publication 3/3.
07AFR Sufetula: Church ‘Basilica 4’ was erected in a street adjacent to the forum, reusing the previous street portico: N. Duval, Les basiliques de Sbeitla à deux sanctuaires opposés (basilique I, II et IV) (BEFAR 118) (Rome 1971) 325–80. It sees a sequence of two churches. I am concerned to date both the first and second church, as an example of ecclesiastical building operations respecting a forum. I limit my remarks to the dating evidence.
The first church is built on a cindery layer, 30–40 cm thick, found inside and outside the basilica. This fill contains the following finds: one coin found inside the basilica that seems to be of Constantius II Fel. temp. reparatio (given as ‘towards 360’); a second coin found outside the basilica (given as ‘late 4th c.?’); unspecified ceramics (of ARS) suggest that the fill does not date before the end of the 4th c. See p. 373 plus p. 344 + p. 347 + p. 385 for the coin of Constantius (no. 1), and p. 373 plus p. 340 + p. 385 for the other coin (no. 3), which seems less significant as found on the surface.
The second church has a floor that has produced a ‘Vandal’ coin [so 439–533] from inside its mortar layer, whilst on the floor itself was recovered a bronze coin of the 13th year of Justinian (p. 373 and 385 no. C5). One tomb in the floor of the second church (no. 6, with inscription IV.3 of the corpus N. Duval “Inscriptions byzantines de Sbeitla (Tunisie), III”, MEFRA 83.2 (1971) (423–43) 436–37 has an epitaph for a Gaudibundus, who died when 9 months old (pp. 379, 385, 373). This inscription Duval would like to be 6th c., because it is in the same hand as a series of epitaphs of children found in Basilica VI, which are dated to 543, as well as the epitaph of Fortunata and the inscription of Matrona dated to 546, citing N. Duval, “Nouvelles recherches d’archéologie et d’épigraphie chrétiennes à Sufetula (Byzacène)”, MEFRA 68.1 (1956) (247–98) 277–81 (nos. 1–5) and 282–83 (no. 7) and pl. 5 (all dated by indictions and the regnal years of Justinian, as explained on pp. 287–94). The second church also has a mosaic that Duval (359–69) would like to date to the 6th c., based on the style of its motifs, of which only parallels to S. Vitale in Ravenna and the ‘Justinianic’ Basilica of Sabratha seem to be securely dated.
A third phase sees a last floor of mortared debris over the top of the mosaic by the central door (perhaps the same mosaic as mentioned above), which seems to be slightly later than the second church. This mortared debris contains a Vandal coin (p. 373).
Overall, I will not use the mosaic style as evidence here. Rather, I offer dating for the first church based on a TPQ of 360, from the coin from the fill. I offer dating for the second church based on a TPQ of 439 from the coin from the mortar, plus a TAQ of the 6th c. from the epitaph from the tomb within its floor. This seems reasonable, given the date of its parallels. The other associated coins I will not rely on here, as only one is clear, from the occupation of the second church, and so does not help us refine its construction chronology further.
Dating summary (first church): range 360–600, midpoint 480, class Cs7 (TPQ coin), Cs6 (TAQ, inscription in situ), publication 3/3.
Dating summary (second church): range 439–600, midpoint 519.5, class Cs7 (TPQ coin), 5 (TAQ, inscription in situ), publication 3/3.
07AFR Sabratha: A church was installed in the civil basilica on the forum. This followed the reconstruction of the civil basilica (without indications of any Christianisation) dated to “soon after the middle of the 4th c.”. The site has been last excavated by J.B. Ward-Perkins and written up by P.M. Kenrick, Excavations at Sabratha 1948–1951 (JRS monograph no. 2) (London 1986). It has recently been restudied from a site visit by A. Leone, The End of the Pagan City: Religion, Economy, and Urbanism in Late Antique North Africa (Oxford and New York 2013) appendix 2. Kenrick had not himself been able to visit the structure for the writing of his report, as Leone records, via a pers. comm noted in her text with her n. 2.
Of phasing the following account can be given:
(i) The reconstruction of the structure as a double-apsed secular basilica, in the style of the Severan basilica of Lepcis Magna, seems to have been part of a general restoration of the forum with much spolia after a traumatic event (war or earthquakes) in 364–75, for which see appendix L1 and K4a. Leone (2013) appendix provides a list of the spolia and its suggested origin, based on the taking of measurements, observing that half of the columns in the basilica-church are of the same size (1.61–1.68 m) as those from the Antonine and Southern temple. The remaining columns come from unknown buildings. Other spolia, such as capitals and decorated pilasters, are suggested to have come from the Early Imperial civil basilica itself. Regardless, the complete rebuilding of the interior of the basilica, to serve the same function, strongly suggests it was part of contiguous later 4th to earlier 5th c. building works in the forum. Leone (2013) appendix notes that the stylobates, which run the whole length of the hall, relate to the civil basilica and not any subsequent churches, as Kenrick (p. 85, fig. 34) thought. These stylobates cut the foundation of a large square base in the centre of the basilica nave, which later served for the altar of the church: Kenrick (1986) 85. Kenrick notes here a coin of 347–48 from its foundation trench, but disregards it as intrusive.
(ii) According to Kenrick, the first definite church involved the installation of an altar in the basilica nave beneath a canopy, supported by reused columns, and the installation of a baptistery, behind the western apse. Leone (2013) appendix notes, from site observations, further features relating to the church, which allows a better understanding of the overall phasing of the structure: the church’s initial colonnade was established on quadrangular foundations that cut into the north stylobate, but which did not extend more than about two-thirds of the way along the length of the basilical hall, unlike the stylobate. Rather, the remainder of the space to the east never had a colonnade, although it did have stylobates, which must reflect the organisation of the civil basilica followed by the first church. Thus, she attributes to this first church phase a north-south wall that divided off the rest of the nave and suggests this part of the hall was kept open to the sky, to serve as an atrium, as suggested by the presence of a drainage system in this area. She and Kenrick place the altar with canopy in the second phase of the church (without stating why), although she accepts that the baptistery belongs in the first phase, perhaps because it was filled and sealed, and replaced by an altar under a canopy of 4 columns, an action that is best situated in the second phase of the church.
(iii) The second church was built sometime after the first was partly ruined and a deep soil layer had accumulated. This involved rebuilding the church to a slimmer design, with a probably new north wall (noted by Kenrick and Leone) and also a new north stylobate, observed by Leone, which is clearly secondary to the quadrangular bases of the colonnade and the related east wall of the church, which should thus be associated with the first church, as they are themselves posterior to the late civil basilica: Kenrick (1986) 85–87 plus Leone (2013) appendix. The colonnades were rebuilt with twin columns, likely obtained from a temple, along with other spolia (including capitals used upside down as bases). A screened area was also created inside the nave: Kenrick (1986) 85–87. See Kenrick and Leone for other details of these churches.
Of dating, we have the following pointers, based on wider theories of site development as much as evidence from the complex itself: (i) The late civil basilica, given that it represents a total rebuild for the same function, is best given the same range of dating as the rest of the work on the forum, as we seem to have a contiguous area of late rebuild for civic secular monuments. Evidence for this renewal begins with epigraphy of 364–75 and has as its latest evidence a ceramic sherd of form Hayes 76 [find no. vi.d.i.1/8], attributed to the “middle quarters of the 5th c.” [LRP/SLRP has 400–475, though form 76a is given as 425–75 and 76b has 400–50] from beneath the mosaic from the portico that unites the basilica to the curia: Kenrick (1986) 33. There is no convincing dating from within the civil basilica, except its use of spolia from elsewhere in the forum, which is problematic, as much of this probably arrived when the second phase of the subsequent church was built. (ii) The first church can most securely be dated by the termini of the civil basilica and the second church. Suggestively, a Tripolitanian lamp of 5th c. date has apparently been recovered from under the floor of the western apse: Kenrick (1986) 82 does not think any work in the apse floor relates to the church phase, but rather places it all in its civil basilica phase, and thus implies it is intrusive (it is “close to the surface”). However, it would be plausible to see it as a TPQ for work on the floor relating to the church, given Leone’s critique of the excavation phasing, which is the interpretation I adopt here, considering it to provide the basis for a ‘poor’ date. (iii) The second church construction likely dates during the reconquest period (i.e. after AD 533 for this region), given that it represents a major restatement of confidence in an otherwise deteriorating site. The building work is dated by its ‘sacral association’ with a cemetery established over the plaza, which has been ascribed to the Byzantine period, based on funerary inscriptions, a grave containing a coin of Justinian or Maurice Tiberius, and also pottery (Kenrick (1986) 86–87). Kenrick notes (p. 34) that the pottery from the backfill of two graves included as their latest pottery “rather more of the latest African forms of the late sixth and seventh centuries (Hayes forms 91D, 105 and 106).” These forms are dated respectively to AD 580–675, 580–675 and 600–75 in LRP/SLRP.
Overall, the initial conversion of the civil basilica to a church is likely to date to sometime in the 5th c. The first church building must date after 400, the apparent start date of the Tripolitanian lamp [if this is correct], and it should be abandoned by ca. 500, given its ruinous state in the overall scheme of development. The second church could be given a ‘Byzantine’ start date of 533, given that the church is indeed a ‘statement’ built in the heart of the city centre, and the surrounding associated evidence dates no earlier than ca. 580. I take the end date for the building of the second ‘Byzantine’ church as 25 years after the start date of the latest ceramics derived from the graves, so a date of 625. All the secular repairs to the basilica as a civil structure ought to predate ca. 500, given the decay of the structure, but we cannot know exactly when in the 5th c. the first church was installed. Kenrick (1986) 84 himself suggested the first half of the 5th c., on ‘historical grounds’, without specifying why.
Dating summary (for conversion of civil basilica into a church): range 400–500, midpoint 450, class Cs7 (TPQ ceramic), Cs3 (associative, phase of development), publication 2/3. Poor.
Dating summary (for construction of second ‘Justinianic’ church): range 533–625, midpoint 579, class z (regional development), Cs6 (absolute, inscription in situ), Cs9 (contextual ceramics), publication 3/3.
07AFR Lepcis Magna (Old Forum): A church was built inside a former temple on the forum: D.E.L. Haynes, An Archaeological and Historical Guide to the Pre-Islamic Antiquities of Tripolitania (London 1955) 86–88; P. Romanelli, “La basilica cristiana nell’Africa settentrionale italiana”, in Atti del IV Congresso Internazionale di Archeologia Cristiana, Città del Vaticano 16–22 ottobre 1938 (Studi di Antichità Cristiana 16) vol. 1 (Rome 1940) (245–89) 266 (not seen); R. Bartoccini, “Una chiesa cristiana nel vecchio foro di Lepcis”, RACrist 8 (1931) 23–52; R.G. Goodchild and J.B. Ward-Perkins, “The Christian antiquities of Tripolitania”, Archaeologia 95 (1953) (1–82) 24–29. The most recent discussion is M. Pentricci, “L’attività edilizia a Lepcis Magna tra l’età tetrarchica e il V secolo: una messa a punto”, in Leptis Magna. Una città e le sue inscrizioni in epoca tardoromana, edd. I. Tantillo and F. Bigi (Cassino 2010) (97–171) 142–43 with p. 139 fig. 4.30 for a plan and p. 140 fig. 4.31 for a photo. Here I limit my observations to phasing and dating, as the date of the foundation of the church is all that interests me here.
Of phasing and dating, we can see a number of arguments emerge. (i) Firstly, we can note the theory that the violet-granite columns for the church came from the adjacent porticus triparta, built on the site of the civil basilica with ‘Troadesian’ (i.e. violet granite) columns in 324–26 (see appendix X1a for inscription and structural details), so giving a TPQ for work on the church: Pentricci (2010) 139 with n. 246 and p. 140 fig. 4.31. (ii) Secondly, Goodchild and Ward-Perkins note that the adjacent ‘Justinianic’ baptistery on the forum, which must be secondary to the church given its dependency on the building, is different in its mortar, which is shelly, like mortar from other ‘Byzantine’ constructions at Lepcis. Pentricci (2010) 142 also notes that the mortar of the church is different to that of the Byzantine fortification wall of 533 to 543, dated based on texts (see appendix V4c). (iii) Thirdly, an ambo foundation has been identified in the church by Tantillo from a large reused block placed in the nave that does not make any sense to have been moved here, given its size, in another structural role. The presence of such a feature suggests work after the Byzantine reconquest of 533, as the ambo as a liturgical element was apparently not well understood in Tripolitania (pp. 142–43). I do not wish to credit this argument, as the ambo might well have been added in a secondary phase. (iv) Fourthly, the inscriptions of the cemetery have been variously dated to the end of the 5th to the beginning of the 6th (Bartoccini) or 6th c. (Duval (1975) below), referring especially to IRT 834 (in the 12th indiction), which could potentially give us a not-very-certain TAQ of the early 6th c. or the entire 6th c. for the church, as one would perhaps not expect to find burials in the city centre before a church had been built there: Pentricci (2010) 142–43 with references. N.B. in N. Duval, Recherches archéologiques à Haïdra: Vol. 1, les inscriptions chrétiennes (CEFR 18) (Rome 1975) 484 with n. 6, Duval notes that the use of an indiction is clear evidence of a Byzantine date in Africa. (v) Fifthly, there are parallels between this church and the second church installed in the civil basilica of Sabratha, dated to 533–625 (see above). According to F. Teichner, “Signa venerandae christianae religionis: on the conversion of pagan sanctuaries in the dioceses of Africa and Aegyptus”, LibSt 27 (1996) 53–66, some structural parallels have been made between the two churches. I was only able to identify one (important) common feature shared by the two structures, namely the aisle colonnades, which are composed of twinned columns, set in a three-aisled plan: J. Christern, “Die Gerichtsbasilica beim Forum von Tipasa”, in Studien zur spätantiken und byzantinischen Kunst. Festschrift Fr. W. Deichmann (Bonn 1986) (163–204) 183–84 with 189, fig. 12b; J. Christern, “Gruppe der zentralisierenden Basiliken mit Vierung des 6. Jh.”, BZ 62 (1969) 287–90. The Sabratha basilica is of a Justinianic date, which depends on the association of a cemetery with the church, spreading over the forum, in which funerary inscriptions and also pottery suggest a ‘Byzantine’ date, whilst one grave contained a coin of Justinian or Maurice Tiberius. The Sabratha church’s association with reconquest building depends on a visible ‘gap’ in the development of the site, in which a layer of soil accumulated, prior to the construction of the church. This suggests a ‘renewal’ in the centre of the city, which would suit the context of the reconquest of the city in 533. See above for details. At both Sabratha and Lepcis, the twinned columns are reused, but their arrangement as double columns is (at least at Lepcis) a new feature: they were not displayed as double columns prior to being incorporated in the church. This is clear from the photo on p. 140 fig. 4.31. At Sabratha, this double column arrangement was not a feature of the stylobate of the first church, but definitely a feature of the second ‘Justinianic’ church, as described by A. Leone, The End of the Pagan City: Religion, Economy, and Urbanism in Late Antique North Africa (Oxford and New York 2013) appendix 2 with figs. 42–43 (relevant phase plans). (vi) Finally, Goodchild and Ward-Perkins believe that the ‘elaborate’ architecture of the church could not be built in the later Vandal years [an argument I do not credit].
Overall, it is not helpful or necessary to adopt all of the arguments set out here. The mortar observations are persuasive: because the baptistery is secondary to the church, we can suggest a TAQ for the church of 543, the upper limit for the construction date range of the fortifications. The parallel with the second church of Sabratha is an associative local one with a neighbouring city, but can be accepted, given that the monuments are both of the same high-status, being churches on fora. Thus, a reconquest date can be accepted, on grounds of architectural style with the better-dated second church of Sabratha, giving us a TPQ of 533.
Dating summary: range 533–43, midpoint 538, class Cs5 (catch-all, masonry), Cs1 (architectural style), publication 2/3. Poor.
07AFR Lepcis Magna (Severan Forum): The civil basilica was converted into a church. This was done through the addition of a raised platform for the sanctuary area, a pulpit, and a vaulted baptistery, alongside various other secondary modifications, such as the removal of decorative architectural elements from the apse, where steps were added for the clergy to sit on: R.G. Goodchild and J.B. Ward-Perkins, “The Christian antiquities of Tripolitania”, Archaeologia 95 (1953) (1–82) 22–24. There are also minor changes within the nave of the basilica, apart from the chancel platform. A reused slab marking the place of the altar and two stairs in reused material were added (J.B. Ward-Perkins “Excavations in the Severan Basilica at Lepcis Magna, 1951”, PBSR 20 (1952) 119–21), quite apart from liturgical ornaments such as the pulpit and chancel screen which included reused parts of the Severan arch (p. 120 with n. 24).
Of architectural form, it has been noted that the civil basilica was structurally intact in its complex internal colonnades, as archive drawings show, at the time of its conversion into a church: M. Pentricci, “L’attività edilizia a Lepcis Magna tra l’età tetrarchica e il V secolo: una messa a punto”, in Leptis Magna. Una città e le sue inscrizioni in epoca tardoromana, edd. I. Tantillo and F. Bigi (Cassino 2010) (97–171) 159. The floor of the civil basilica had been stripped of its Proconnesian marble before a new fill and a higher floor was laid down for the chancel area: Ward-Perkins (1952) (111–21) 111. The southernmost of the 4 chapels that filled the angles of the building also showed a raised new floor, but one which reused the original slabs, at a higher level, which must have survived in situ until this moment. In the west-corner chapel (converted into a baptistery) the upper ecclesiastical floor sealed two angle columns of cipollino marble, showing that at least some of the basilica’s upper parts were in decay at the time of the conversion into a church. In this chapel and in the south and east chapels the roof was replaced by a barrel vault: Ward-Perkins (1952) 111. Some of the original basilica paving was also reused to repave the baptistery and to pave a room leading into the colonnaded street, which had been raised up with a new floor. The former room was in its second late phase of modification, its exit to the colonnaded street having been blocked and low benches installed on one side of it, which led Ward-Perkins to suggest it had been made into a synagogue: Pentricci (2010) 159; Ward-Perkins (1952) 111, 112–19.
Of dating, this conversion into a church is supposed, by Goodchild and Ward-Perkins (1953) 22–24 and by Duval (below) to have taken place under Justinian: it is identified with a church dedicated to the Virgin Theotokos under the emperor, mentioned by Procopius: N. Duval, Les églises Africaines à deux absides, vol. 2 (BEFAR 118bis) (Paris 1973) 279–82, with previous bibliography; Procop. Aed. 6.4.4–5. If we take Procopius’ text as applying to this building, the dedication must date after the Eastern reconquest in 533, but before 558, the end date for his panegyric. The end-date is derived from the collapse of the first dome of Hagia Sophia. This collapse is not anticipated in Procopius’ text of the Aed., despite a description of Hagia Sophia in 1.1.66–78, which discusses the stability of the dome, a lacuna which would be odd, given the expectations of the reader after 558. However, there is nothing on site or in Procopius’ description to confirm the identification of the church. One could note that it is the only dedication described by Procopius in Lepcis, and that our church is of great size and prominence. These factors make Procopius’ church a reasonable identification for the church in the Severan Forum.
Dating summary (conversion into a church): range 533–58, midpoint 545.5, class x (historical texts), Cs3 (associative, phase of development), publication 3/3.
See also (appendix S5a): 02HIS Ilici: Church built inside forum temple, sometime in 575–725.
U3 Churches on but Not over Fora / Agorai 4th–5th C.: the East
12CPL Constantinople (Augusteion): The Great Church was begun around the time that bishop Paul was sent into exile: Socrates, Hist. eccl. 2.16.16. This happened under praetorian prefect PLRE 1.696–97 Fl. Philippus 7, who was in office ca. 344–51. Paul died shortly after the emperor Constans, so probably in 350: Sozom., Hist. eccl. 4.2. The church was dedicated in 360: Socrates, Hist. eccl. 2.43; Sozom., Hist. eccl. 4.26 (both giving consular dates). This first church of Hagia Sophia has not been uncovered in Istanbul, but the position of the building, adjacent to the Augusteion square, is known from contemporary textual descriptions, before it was destroyed in 404: (i) Shops on the south side of Hagia Sophia, probably on the Augusteion, were removed to build a monastery around 400: Vita Olympiadis 6; (ii) In the Augusteion a silver statue monument (appendix H9) to the Empress Eudoxia was erected in 403, which drew the anger of John Chrysostom, as the dedication ceremony could be heard inside the adjacent Great Church: Socrates, Hist. eccl. 6.18 (locating the statue one half street breadth away from the Church of St. Sophia); Socrates’ re-writer Sozom., Hist. eccl. 8.20, states that it was opposite the Senate House.
Dating summary (for dedication of church): range 360, class x (historical text), publication 3/3.
15ORI Jerusalem (‘Forum’): The Temple of Venus-Aphrodite was demolished to build the Holy Sepulchre in AD 326: Euseb., Vit. Const. 3.25–43. The position of a forum adjacent to this church is highly controversial. The church, as we know its ruins, is awkwardly orientated to face west, opening east towards the street rather than to the north or south, where a forum might be (see appendix E3). The argument that it was set on a forum seems to depend on Egeria 43.7, who records that its gates opened onto the market (quintana). This could of course mean the main street, which we know from excavation that it opened onto, rather than onto any adjacent forum. I have never seen the word quintana used to mean forum in any other late antique text.
Dating summary: 326, class x (historical text), publication 3/3.
16AEG Alexandria (Augustan Agora): The Kaisarion was given by Constantius II as the site for the Great Church between AD 339 and 345 (under the episcopate of Gregory of Cappadocia): Epiph. adv. Haer. 69.3.3 (GCS 37 p. 153). This building may have stood on one side of a forum, but the theory is far from settled, depending on suggestive Early Imperial evidence: J. McKenzie, The Architecture of Alexandria and Egypt 300 BC–AD 700 (New Haven, Connecticut-London 2007) 177. For mention of the ‘Σεβα(στὴν) Ἀγορά(ν)’ (‘Augustan Agora’), probably located at or inside the Kaisarion, see BGU 4.1079 (http://papyri.info/ddbdp/bgu;4;1079 (last accessed July 2016)) (AD 41). See also Apthonii Progymnasmata 12, ed. H. Rabe (Leipzig 1926) p. 38 for location as south of the Caesareum, on the north side of the main east-west street.
Dating summary: range 339–45, midpoint 342, class x (historical), publication 3/3.
See also (appendix A1): 11THR Novae: A large church with baptistery, dating sometime in 395–616, was built adjacent to a ‘forum’ (courtyard of former principia building).
See also (appendix A4b): 10MAC Philippi: Basilica B, just south of the forum, hides an earlier basilica church, detected by N. Karydis, which is very similar to Basilica A of the same city of 450–527, although it cannot itself be more closely dated than 313–563.
U4 Churches on but Not over Fora / Agorai, 6th–7th C.: the East
13ASI Sagalassos (Upper Agora): The bouleuterion courtyard, on the west side of the Upper Agora, was occupied by a church. This has recently been re-dated to the 6th c. The church was linked to the agora by a staircase built up to the former bouleuterion itself, which now served as an atrium for the church: see appendix S6. The re-dating is the result of sondages that revealed: (i) ceramics of 6th c. date from under paving, which was disturbed to install a water channel for the church; (ii) ceramics from the second and third quarters of the 6th c., found in the levelling subsurface for the mosaic floor of the church: P. Talloen and J. Poblome, “Control excavations in the bouleuterion basilica”, Anmed 2016 (119–128) 125–26. The latter ceramic information provides a closer range of dating than is usual for Sagalassos Red Slip Ware. However, as with other ceramic dates in the city, we have a wide date range given for a type of ceramics, rather than knowing the start dates of single wares. This means we have to give a dating of the full range given for the last pottery, which is a relatively crude dating measure.
Dating summary: range 525–75, midpoint 550, class Cs9 (contextual pottery), Cs3 (associative, phase of development) publication 2/3.
Discounted: 15ORI Zenobia: As noted in the chapter on 6th c. agorai, recent excavations have confirmed that the ‘agora’ at Zenobia is not really an agora. Therefore, I will not discuss the two adjacent churches. See S. Blétry, “Trois années de recherches à Zénobia-Halabiyé (Syrie), ville forteresse proto-byzantine sur le limes oriental”, Semitica et Classica 3 (2010) (249–64) 249 and 258–64; S. Blétry, “Zénobia Halabiyé. Campagnes 2007”, http://recherche.univ-montp3.fr/cercam/article.php3?id_article=497 (last accessed February 2017).
U5 Churches on, but Not over, Agorai, Unspecified Date in Late Antiquity
05ITA Trieste: The church next to the forum was built inside the propylon of the capitolium, in reused materials, at an unknown time during Late Antiquity. This church was restored / improved by bishop PCBE 2.1.876 Frugifer (known around AD 542 and 565), as recorded in an inscription in a mosaic floor within the church: M. Mirabella Roberti, “Considerazioni sulla basilica suburbana di Trieste”, Atti dei Civici Musei di storia ed arte di Trieste 6 (1969–70) 101–12; V. Scrinari, Tergeste (Trieste): Regio X, Venetia et Histria (Italia Romana: Municipi e Colonie 1.10) (Rome 1951) 97–98; M. Mirabella Roberti, “Mosaici paleocristiani scoperti negli scavi della basilica S. Giusto”, in Giornale di Trieste (5th March 1949). For the inscription, see A. Degrassi ed., Inscriptiones Italiae vol. 10 Regio X part 3 Histria Septemtrionalis (Rome 1936) 168. For the mosaic inscription and the bishop’s monogram recorded on capitals on this church, see A. Degrassi and P. Sticotti edd., Inscriptiones Italiae vol. 10 Regio X part 4: Tergeste (Rome 1951) no. 295 (mosaic inscription with plan of mosaic floor) and nos. 296–97 (monogram on two capitals). There is nothing to provide a TPQ for the initial construction of the church, other than the Edict of Milan in 313. We cannot be certain that 565 is the last possible date for Frugifer, but 565 is a reasonable TAQ indication of the latest date for the church. On the original structure of the propylon, see C. Zaccaria, “Problemi epigrafici del foro di Trieste”, MEFRA 100.1 (1988) (63–85) 67–70.
Dating summary: range 313–565, midpoint 439, class x (historical text), Cs6 (TAQ inscription in situ), publication 3/3. Generic late antique date.
08PAN Zadar in Dalmatia: The first church was built within a row of three shops on the forum. The second phase is thought to be of the first half or middle of the 5th c., based on mosaics. The shops were thought by P. Chevalier and others to have been abandoned in the late 4th or early 5th c., but I have been unable to obtain dating evidence: P. Chevalier, Ecclesiae Dalmatiae: l’architecture paléochrétienne de la province romaine de Dalmatie (IVe–VIIe s.): en dehors de la capitale, Salona (Split 1996) vol. 1: 100–107, vol. 2: plates 28–29.
09DAC Serdica: The ‘St. George Complex’ near the forum seems to be an incomplete bath building that was converted into a church. For a bibliography and references to undated theories on church conversion, see L. Lavan, “The residences of late antique governors: a gazetteer”, AnTard 7 (1999) (135–64) 144–45.
10MAC Thasos: A church was built alongside the agora, in two phases, which incorporated bits of classical buildings: AA.VV., “Chronique des fouilles et découvertes archéologiques en Grèce en 1950: Thasos”, BCH 75 (1951) (142–87) 154–64. It is thought to be 5th c. in its first phase, the dating of which (p. 164) is based on mosaic style (pp. 160–61, no reasons given), the forms of letters of a dedicatory inscription (p. 158 where given as 5th–6th c.) and the style of the jewelled crosses painted on the walls (p. 158, known from the second half of the 4th c. in imitation of the jewelled cross then displayed at Golgotha in Jerusalem). An earlier building was detected on the site, thought to be 4th c., on the basis of chancel screen fragments from this period (p. 158 with p. 163). Overall, I can accept that the dedicatory inscription suggests a date in the 5th–6th c. for the second church; I do not think the style of the cross enjoys enough dated parallels to be used here. It is more difficult to find a date for the first ‘church’, other than a TPQ of the Edict of Milan in 313, after which we can envisage that monumental church building was tolerated. This TPQ is not advanced by the presence of a 3rd c. inscription, reused in the construction of the crypt, nor by other reused material. The only TAQ for the ‘first church’ is 600, derived from the upper limit of the dedicatory inscription of the second church. In support of dating, one can note that there was an urban crisis that affected Thasos in the early 7th c., visible in a late antique domus on the site: see appendix A8b.
Dating summary (first ‘church’): range 313–600, midpoint 456.5, class x (historical text), Cs3 (associative, phase of development), Cs6 (inscription, in situ), publication 3/3. Generic late antique date.
13ASI Rhodes: A basilica church, ca. 70 m long, was built along the north side of the agora: E. Kollias, “Η παλαιοχριστιανική και βυζαντινή Ρόδος. Η αντίσταση µιας ελληνιστικής πόλης”, Ρόδος 2400 χρόνια. Η πόλη της Ρόδου από την ίδρυσή της µέχρι την κατάληψη από τους Τούρκους (1523) (Athens 2000) 299–308 (not seen). I thank G. Deligiannakis for this reference.
13ASI Xanthos (Lycia) (Theatre Square / Roman agora): A church on the west side of this square was not excavated, but is described in J. des Courtils and D. Laroche, “Xanthos—Le Letoon. Rapport sur la campagne de 1998”, Anatolia Antiqua 7 (1999) (367–99) 373–76.
13ASI Antiphellos (Kaş): A church built on one side of the agora is present on the map of the town by Texier (second half of the 19th c.), but is no longer visible today: C.V. Daremberg and E. Saglio, Dictionnaire des antiquités grecques et romaines vol. 1 (Paris 1872) 152 (with thanks to P. Talloen). There are, however, some late antique capitals, which likely come from the church and are displayed today at the nearby ‘Building of the District Authorities’, (P. Talloen pers. comm., town resident and archaeologist), when compared with examples in T. Zollt, Kapitellplastik Konstantinopels vom 4. bis 6. Jh. (Asia Minor Studien 14) (Bonn 1994). The volutes of the capitals from Antiphellos resemble a later 4th to earlier 5th c. example from Constantinople (e.g. no. 303); its leaf stems resemble a later 5th c. example (e.g. no. 391); its pointed leaves resemble some later 5th to first half of 6th c. examples (e.g. no. 617), which also resemble it in the design of its lower leaves. However, no capital from Constantinople is the same as that from Antiphellos. The level of doubt is such (none of the Constantinopolitan parallels seem to be firmly dated) as to render the date of the church as anytime from 350–550. For a photo of the capital see http://www.ajanstr7.com/files/news/default/2016/01/chp-kas-tan-kaymakamlik-ziyareti.jpg (last accessed Jan 2016).
Dating summary: range 350–550, midpoint 450, class Cs1 (architectural style), publication 3/3.
13ASI Cremna: The conversion of the civil basilica into a basilical church occurred sometime in Late Antiquity. It is documented, with few details, apart from basic observations on the church plan, by S. Mitchell et al., Cremna in Pisidia: An Ancient City in Peace and in War (Swansea 1995) 221–22. See also J.B. Ward Perkins, M.H. Ballance and J.M. Reynolds, “The Caesareum at Cyrene and the basilica at Cremna, with a note on the inscriptions of the Caesareum by J.M. Reynolds”, BSR 26 (1958) (137–194) 173. Under the rules of this study, we can only give it a generic late antique date, based on the construction being after the Edict of Milan in 313, but before the decisive Persian invasion of 616, after which the economic downturn in the region would make such a large building project unlikely, except in the case of fortifications.
15ORI Beroia (Aleppo): The bishop’s church (not dated, though suggested as second half of 6th c.) stood on / near the ‘agora’. In AD 715 a large mosque was built over the ‘agora’. Sauvaget argued that an agora existed on this spot, because of its central position because of a gap in the colonnades of the main avenues, and because aqueducts reached a terminus at this point: J. Sauvaget, Alep: essai sur le développement d’une grande ville syrienne des origines au milieu du XIXe siècle (Paris 1941) 47, 59, 75. He cites Ibn Shaddād 37, Pearls 56, not seen by me, for the date of the foundation.
Dating summary: sometime prior to 715, class x (historical text?), publication 3/3. Generic late antique date. Poor, as I have not been able to check the sources.
15ORI Samaria / Sebaste: The conversion of a civil basilica into a church may have occurred in Late Antiquity. Two apses were added at the north end, a smaller one inside the other, resembling the synthronon of a church. They are, however, not dated more precisely than to Late Antiquity, based on the reused material included in their masonry. This rebuilding could have occurred in the Umayyad period also, or even a bit later: J.W. Crowfoot, K.M. Kenyon and E.L. Sukenik, The Buildings at Samaria (Samaria-Sebaste 1) (London 1942) 37.
See also (appendix K4c): 10MAC Elis: A church was built in the ruined stoa of the agora, sometime 313–616, a generic late antique date. 11MAC Tegea: A church was built in a stoa. I was unable to access a publication on this, so no dating can be given.
See also (appendix L3): 13ASI Selge: A church was built inside the bouleuterion, at an undated time within a generic late antique range of 313–614.
Discounted: 06ITS Terracina: The cathedral is located on the present forum, but no archaeological work has placed a building on this site in Late Antiquity. The presence of a later 6th to 7th c. inscription on a column in the porch in front of the church, recording the restoration of the forum, has led to suggestions that the church also dated from this time. However, the column could well have been spolia from the forum, reused in the church; 13ASI Magnesia on the Meander (no church conversion recognised in recent survey) and 15ORI Tiberias (site recently identified as the seat of the Sanhedrin, does not resemble a civil basilica or a church): see appendix X1d (discounted section) for both of these last two entries.
V1 Churches Built over Fora / Agorai, the West (Dated)
01BRI Lincoln: Two possible churches, orientated east- west, were built over the open plaza of the forum, one on top of the other: M.J. Jones, “St. Paul in the Bail, Lincoln. Britain in Europe?”, in Churches Built in Ancient Times. Recent Studies in Early Christian Archaeology, ed. K. Painter (Society of Antiquaries: Occasional Papers 13) (London 1994) 325–47.
Of phasing, the first structure is a rectangular two-room structure, oriented east-west, of which the eastern room, being around 4 times smaller than the western room, is mimicking the organisation of a church with nave and apse. The second structure is an apsidal hall, also oriented east-west. These buildings are best identified as churches as a default interpretation, given their orientation and form, despite no sacred art or inscriptions being recovered from the site, even if this basis is flimsy compared to Mediterranean sites. I do not think it is clear what the relationship of the church is to the demolition of the forum ca. 400, discussed in appendix K4d. Some artistic reconstructions show the church accessed from a still-standing portico of the forum, but the plans show this area was not excavated, with the church standing in the very middle of the open plaza: e.g. M.J. Jones, “The latter days of Roman Lincoln”, in Pre-Viking Lindsey (Lincoln Archaeological Studies 1), ed. A. Vince (Lindsey 1993) (14–28) 26 fig. 3.9 (plan with excavation boundaries).
Of dating, the first structure is given a TPQ only by a single coin of Arcadius (AD 388–92) (on p. 329). The second structure is cut by three burials, which have produced radiocarbon dates at 95.4% confidence level (2 standard deviations) as follows: (i) The first is of cal. AD 419–765 (HAR 4131 / Burial 519 / Sample 30 of CG 250 / LUB 32); (ii) The second is of cal. AD 433–775 (HAR 4116 / Burial 517 / Sample 29 of CG 262 / LUB 32); (iii) The third is of cal. AD 350–687 (HAR 5092 / Burial 507 / Sample 26 of CG 266 / LUB 32): K. Steane, Archaeology of the Upper City (Lincoln Archaeology Studies 1) (Lincoln 2006) 156–58, 161, 210 refined by G. Gilmour, “Sub-Roman or Saxon, pagan or Christian: who was buried in the early cemetery at St. Paul-in-the-Bail, Lincoln?”, in Pagans and Christians: from Antiquity to the Middle Ages: Papers in Honour of Martin Henig, Presented on the Occasion of his 65th Birthday, ed. L.A. Gilmour (BAR-IS 1610) (Oxford 2007) (229–58) 246–49 (taking the figures from the latter publication). A Bayesian statistical analysis suggests an >85% chance that they were in place by 600 [which I will take as the TAQ for the church]: T. Green, Britons and Anglo-Saxons: Lincolnshire AD 400–650 (Studies in the History of Lincolnshire 3) (Lincoln 2012) 65–69, from which I derived references here.
Dating summary: range 388–600, midpoint 494, class Cs7 (TPQ coin), Cs10 (scientific, radiocarbon), publication 3/3.
04VIE Aix: A baptistery was built over the forum. For the main excavation reports, see R. Guild, J. Guyon and L. River, “Recherches archéologique dans le cloître Saint-Sauveur: bilan de quatre campagnes de fouilles (1976–79)”, RANarb 13 (1980) 115–64; R. Guild, J. Guyon and L. Rivet, “Les origines du baptistère de la cathédrale Saint-Sauveur. Etude de topographie aixoise”, RANarb 16 (1983) 171–209, with J.-P. Pelletier et al., “Annexes”, in the same volume 210–32; M. Gauthier, “Circonscription de Provence-Alpes-Côte-d’Azur”, Gallia 44.2 (1986) (375–483) 379–83. Selected summaries: J. Guyon and N. Nin, “III: Antiquité tardive”, in Carte archéologique de la Gaule 13/4 - Aix-en-Provence, Pays d’Aix, Val de Durance, edd. F. Mocci and N. Nin (Paris 2006) (413–442) 419–23; R. Guild, “Le cœur antique d’Aix. L’acquis des fouilles archéologiques”, in La Cathédrale Saint-Sauveur d’Aix-en-Provence, edd. P. Coste et al. (Aix-en-Provence, 4th edn. 2008) 11–15 (not seen); R. Guild, J. Guyon and L. Rivet, “Aix-en-Provence: groupe épiscopal Saint-Sauveur-Sainte-Marie”, in Les premiers monuments chrétiens de la France vol. 1: Sud-Est et Corse, ed. N. Duval (Paris 1995) 109–117.
Of phasing, the installation of a baptistery in the north-west corner of the forum, encroaching more than half the width of the plaza, involved raising the ground level considerably, with a massive fill being deposited. This was done so that the baptistery floor reached the same height as the monumental building to the north (a civil basilica or temple perhaps), which is where the present cathedral stands, although there is no trace of the late antique cathedral thanks to 18th c. building works. The deposition of such a fill to reach this height implies that a church already stood on the north side of the plaza in Late Antiquity.
Of dating, the baptistery contains much reused material in its colonnades and stylobate. However, the key dating evidence comes from the fill of the deambulatory floor within the baptistery. This fill has produced ceramics given as dating to 470–90, described as fineware DSP grey or orangey and ARS D, alongside coarsewares with grey fabric. The DSP is described as Rigoir 6b, 8, 9 or 16, 18. All of these DSP forms are given a dating to 370–500 in DCAMNO. The ARS D is described as Hayes 59B [ca. 320–420 in LRP], 60 [ca. 320(?)-380 in LRP], 61A [ca. 325–380 in LRP with SLRP], 67 [ca. 360–470 in LRP], 3A [60–90 LRP], 87A [450–500 LRP] and 87B [500–512.5 LRP], 182 [140–400 in Bon, who divides it into A-D with different date ranges] and a variant of 193 [not dated in LRP, beyond noting of 1st c. BC parallels, although 387.5?–400? is tentatively suggested by Bon, presenting only one example]. Two small bronze coins of the reign of Constantius II were present: Guyon and Nin (2006) 423, drawing on Guild, Guyon and Rivet (1983) 184 and Pelletier et al. (1983) pp. 213–14 (for the ceramics, noting in fact that Hayes 87A and 87B come from the related ‘fill of the cloister’ [i.e. immediately east of the baptistery], alongside 73A [420–75 according to E. Vaccaro pers. comm.] and pp. 229–30 (for the coins, only one of which was strictly in context, coming from the fill layer of the floor of the baptistery). Of these wares, 87B seems the latest of ARS D.
Overall, the evident rigour of the ceramic identifications encourages deference to the local knowledge of the excavators here, and their dating of 470–90. A straightforward contextual date generated under the rules of this study for the deposition of this fill layer is very similar, being 500–525 (based on 87B being the latest find). During this time the construction can be suggested as having taken place. Thus, I will use my own contextual date for the construction of the baptistery.
Of post-church phasing, the late antique level is covered by a layer containing a lot of demolition debris, including marble revetment. This makes one think that the installation of the baptistery was followed by demolition of the forum (revealed in sondage 1, ca. 15 m to the east of the baptistery, for example): Guild, Guyon and Rivet (1980) 116 fig. 2. A cemetery also developed on the site, but it cannot easily be dated, although graves are known. An epitaph fragment, reused in later structures, is believed to date from ca. 500 (132, 154 no. 4).
Dating summary: range 500–525, midpoint 512.5, class Cs9 (contextual ceramics), Cs8 (contextual coins), publication 3/3. Exemplary.
06ITS Herdonia: Between three and 5 ‘chapels’ were built over the forum, apparently during the 5th c. These are well-constructed apsidal structures in brick-faced concrete.
One of these (set inside the civil basilica) was constructed in ‘opus vittatum’, above and later than layers 0.6 m thick, under which the Late Roman occupation was found. The layers 0.6 m thick contained coins of ‘Theodosius and Valentinian’, a bronze brooch, “a balestra” and an African Lamp, alongside architectural fragments from the basilica (including capitals). A second ‘chapel’ in opus listatum (the description sounds like what I call opus vittatum mixtum) was built in a taberna on the eastern angle of the forum, over a layer of 1 m of earth including unspecified 3rd to 4th c. finds. It had a room added later, which incorporated statue bases: J. Mertens, “Dal tardo antico all’altomedioevo”, in Herdonia. Scoperta di una città, ed. J. Mertens (Brussels and Rome 1995) 345–47. On the lamps see C. Delplace, “Présentation de l’ensemble des lampes découvertes de 1962 à 1971”, in Ordona IV, Rapports et Etudes (Brussels and Rome 1974) (7–101) 80 tav. XIX nn. 414–15 (lamp in grey-pink fabric and lamp in ARS with medallion featuring a cross in relief of type ‘Bailey Vict. Alb. Mus. 73 pl. 11, no. 258’ plus other references) On the coins see: S. Scheers and F. Bex, “Les monnaies trouvées durant les campagnes de 1964 à 1968”, in the same volume (105–37) 129 n. 116 (coin of Theodosius I, Arcadius or Honorius) and p. 131 n. 131 (undated coin of 3rd–4th c.).
Unfortunately, there are some problems in accepting Mertens’ statements. The coins do not correspond in the cited report catalogue (although I followed the references) as being the same as quoted in Mertens’ text (no. 119 on p. 126 is of Theodosius II and Valentinian II, though not cited by Mertens). However, the lamps do correspond, and can also be confirmed as correct by the concordance of finds numbers. From this material, the earliest start date of the last confirmed coin is 379, when Theodosius became Augustus, and the rest of the material suggests that it was laid down in a short period. Mertens is clear that chapel 1 is later than these layers, not contemporary with them. As these layers do not seem to represent a direct structural fill, their finds can only give us a TPQ for the chapel itself.
For further discussion of the ‘chapels’, see G. Bertelli, “La trasformazione di un foro romano. Alcune riflessioni sulle cosiddette ‘cappelle’ di Herdonia”, in 1983–1993: dieci anni di archeologia cristiana in Italia. Atti del VII Congresso nazionale di archeologia cristiana. Cassino, 20–24 settembre 1993, vol. 2, ed. E. Russo (Cassino 2003) (733–43) 736 and more recently the more detailed study of D. Leone and G. Savino, “Le ‘Cappelle’ del foro di Herdonia: nuovi dati dall’archivio Mertens”, in 34° Convegno nazionale sulla preistoria—protostoria—storia della Daunia (San Severo 16–17 novembre 2013), ed. A. Gravina (San Severo 2014) 397–412, esp. 399–400. In both cases, the discussions cite the above specialist reports, but the coins are given as of ‘Theodosius’ and ‘Valentinian’ (I or II is not specified), in the manner of Mertens, although Bertelli only mentions a coin of Theodosius I, whilst offering a 3rd to 5th c. date for the lamp in the main text and a 4th–5th c. in a footnote. Leone and Savino (2014) 399–400 confirm that the masonry of both structures is opus vittatum mixtum. I do not accept the view of Leone and Savino (2014) 400 that the finds suggest an absolute date of the second half of the 4th / late 4th to early 5th c.
Overall, the finds from the forum do not seem to provide anything other than a TPQ of 379 for the structures built in the plaza. In the absence of specific seismic damage or evidence of reinforcement on the forum, I am not yet able to accept a generalised earthquake of 346 [Guidoboni Catalogo no. 54] to date the disuse of the forum, even if some reinforcement of adjacent baths buildings has been suggested as relating to this event: G. Volpe, “Herdonia romana, tardoantica e medievale alla luce di scavi recenti”, in Ordona X. Ricerche archeologiche a Herdonia ed. G. Volpe (Bari 2000) (507–54) 523–525. Nonetheless, I do understand that in areas adjacent to the site, in the Baths and the domus, stratigraphy excavated in the last twenty years, shows a degradation of life between the end of the 5th c. and the beginning of the 6th c., in which public monuments are obliterated and humic earth accumulates over collapse deposits, whilst public buildings become quarries for reuse and the site of cemeteries. In the second half of the 5th c., the reoccupation is small-scale and sees perimeter walls reused for structures built of wood and perishable materials. I am grateful to D. Leone pers. comm. for these precisions, though these observations do not seem to have been fully published yet. Thus, I am prepared to tentatively give the more substantial masonry buildings on the forum a TAQ of AD 500.
Dating summary (beginning of disuse): range 379–500, midpoint 439.5, class Cs9 (contextual ceramics), Cs8 (contextual coins), z (site development), publication 2/3.
06ITS Alba Fucens: There is a late antique apse over forum paving, blocking a street completely and reducing the width of another: J. Mertens, “Recenti scavi ad Alba Fucens”, in Il Fucino e le aree limitrofe nell’antichità. Atti del convegno di archeologia, Avezzano (Rome 1991) (387–402) 388. This church includes a statue base in its wall and much reused material alongside tiles without regular coursing. The area is currently under excavation and so may yield some dates soon: L. Lavan site observation, April 2016. As an ecclesiastical structure, it is possible it dates somewhat later than Late Antiquity [the excavators suggest 4th to 8th c.].
Dating summary: Undated.
06ITS Metapontum: A church was built over the Agora and has been tentatively dated to between the mid-4th c. and the end of the 5th c., because at the former date there is a profound rupture in the organisation of the settlement, notably its roadways, into which the construction of the church would fit well, whilst the site seems to have been abandoned in the first half of the 6th c., after a crisis after the end of the 5th c. Its baptistery is compared to examples believe to be 6th c. in Nurachi, Cronus and Tharros in Sardinia and mid-5th c. in San Giusto in Apulia: L. Giardino, “La fascia ionica della Basilicata in età tardo-antico. Continuità e transformazioni”, in L’Italia meridionale tardo antica. Atti del 38 convegno di studi sulla Magna Grecia, Taranto 2–6 ottobre 1998, edd. L. Giardino, R. Auriemma, and E. Lapadula (Taranto 2000) (343–60) 352–53. I take the construction of the church to date between the two wider termini for site development, the grounds for which are not referenced in the article I found.
Dating summary: range 350–500, midpoint 425, class z (site development), publication 1/3.
07AFR Diana Veteranorum (Zana): A church was built over a plaza, which Duval takes to be the forum. Duval notes, from the character of the foundations (irregular and with much reused material in lower courses), that the church was built when the forum was already covered with debris. He thinks a Byzantine date is likely, based on the presence of a fort of this date here. He also expresses his very general impressions of the state of agorai in Byzantine times: N. Duval, “La ‘basilique’ de Zana (Diana Veteranorum)”, MEFRA 89 (1977) 847–73, esp. 855–57 and 869–70. These are, of course, arguments prone to circularity. It is better to say there is no dating evidence so far. The fact that the church contains a good deal of spolia, some of which Duval suggests may come from the forum, means that an analysis of it may well produce a TPQ for the forum in future. There seems to be nothing to date the church, except for the presence of reused material, and a generic dating of sometime ‘after the Edict of Milan in 313’ (which is what I will use here) but before the Muslim invasion of 647. After this date, a church built over a forum is unlikely, given that this is a prime site within the city.
Dating summary: range 313–647, midpoint 480, class Cs4 (reused material), x (historical), publication 3/3. Generic late antique date.
07AFR Lepcis Magna (forum vetus): A square baptistery, with a cross-shaped font at its centre, was built directly over the slabs of the forum vetus, incorporating a Severan statue base. It is thought to be ‘Justinianic’ on account of the shelly mortar of the perimeter wall, which is characteristic of ‘Byzantine’ building work at Lepcis, and on account of the form of the font (no details given): R.G. Goodchild and J.B. Ward-Perkins, “The Christian antiquities of Tripolitania”, Archaeologia 95 (1953) (1–82) 24–29, esp. 27, followed by F. Teichner, “Signa venerandae christianae religionis: on the conversion of pagan sanctuaries in the dioceses of Africa and Aegyptus”, LibSt 27 (1996) (53–66) 56. The baptistery served the ‘Justinianic’ church on the same forum, described above, and was likely secondary to it, given the baptistery’s haphazard provision for what was a very ‘elaborate’ church, with different mortar. The Lepcis church is dated from parallels to the Sabratha church inside the civil basilica; this is dated to 533–625 based on overall site development and historical context, alongside pottery, inscriptions, and coins from the adjacent cemetery, which spread over the forum: see appendix U2 above. Thus, we have a tentative TPQ for the baptistery from the church of 533 onwards, based on architectural parallels to a dated church in Sabratha. Given that the baptistery is associated by its mortar with other (unspecified) ‘Byzantine’ works at Lepcis, we should envisage a date within the reconquest period for the baptistery. A date after the Arab conquest of 647 is unlikely for a prominent Christian monument set in the centre of the city.
Dating summary: range 533–647, midpoint 590, class Cs5 (catch-all masonry: mortar), Cs1 (architectural style), Cs3 (associative, phase of development), Cs9 (contextual ceramics), Cs8 (contextual coins), Cs6 (inscriptions, in situ), z (regional development), publication 2/3. Poor.
V2 Churches Built over Fora / Agorai, the East (Dated)
15ORI Elaiussa Sebaste: The tetragonal agora was converted into a double-apsed church. This development has taken some time to date properly, in the publications of the present Italian team. Dating initially depended upon mosaics from the church, dated to the 5th to the 6th c AD on stylistic grounds, and on the lettering of a funerary inscription found inside the building that has been dated to the 6th c: E. Equini Schneider ed., Elaiussa Sebaste I: campagne di scavo, 1995–1997 (Bibliotheca Archaeologica 24) (Rome 1999) 201–202 (tomb details and finds), 248–50 (mosaic and parallels), 259 (destruction level in trapezoidal room on south side of eastern apse, with coins of reign of Heraclius of AD 613–14) 263 (first mention of tomb inscription of tomb US 46, with dating to 6th / 7th c.); E. Equini Schneider ed., Elaiussa Sebaste II: un porto tra Oriente e Occidente (Bibliotheca archaeologica 37) (Rome 2003) vol. 2.1 p. 202 (tomb US192 with 6th–7th c. dress ornaments, of which I will not go into the details here) 219–22 (discussion of architecture). See also vol. 2.2 pp. 538–40 (discussion of epitaph inscription, given as 6th c., based on palaeography and on the finds cited above). In a recent synthesis, a revised date for the installation of the church is given in the second half of the 5th c., based on the combination of the mosaics, the funerary inscriptions, unspecified ceramics and the architectural ornamentation, notably the capitals, twin columns, pillars and transennae decorated with relief work: E. Equini Schneider, Elaiussa Sebaste. A Port City between East and West. An Archaeological Guide (Istanbul 2008) 54–61 with 61 for the dating.
The appearance of a final report has permitted a clearer understanding of the chronology of the church, which had more than one phase, as well as an archaeological dating of its first phase, based on pottery and coins recovered from a fill relating to the construction of the church. The fill contained a coin of 403–408 and a coin of 408–19 (of Theodosius II), alongside 3rd to 4th c. ceramics. This has suggested to the excavators a date for the initial construction of the church “alla metà del V secolo”: C. Morselli, “La distruzzione dell’agora e l’impianto della basilica protobizantina (Fase 3: metà V sec. d.C.)”, in Elaiussa Sebaste III: L’Agora Romana, ed. E. Equini Schneider (Bibliotheca Archaeologica 37) (Rome 2003) (54–89) 60–61. For the coins, see p. 175 no. 65 (Theodosius II AD 408–419 of context US AP810). For the ceramics, see pp. 193–96, which are mostly residual wares of the 3rd–4th c., with the latest datable wares apparently being Palestinian amphorae of LR9, produced between the 5th and the 7th c., and amphorae from Sinope in the Black Sea, in a variant of the 5th c. [neither of which I have been able to check the dates of]. Under the rules of this study, the latest material provides a contextual date for the building work associated with this fill of 25 years after 408 (the latest start date of a find from these layers). This 25 year period is the standard time gap for construction, following the deposition of a building fill, under the rules of my study. Other chapters in the same book discuss the chronology of later phases of the structure.
Dating summary: range 408–33, midpoint 420.5, class Cs9 (contextual ceramics), Cs8 (contextual coins), publication 3/3.
13ASI Iasos in Caria: Two ecclesiastical buildings were constructed over the agora, one on top of the other: C. Laviosa, “La campagne de Iasos en 1983”, Kazi Sonuçlari Toplantisi 6 (1984) (283–88) 284–85; C. Laviosa, “La campagne de Iasos en 1984”, Kazi Sonuçlari Toplantisi 7 (1985) (293–97) 293–94; F. Berti, “Nouvelle préliminaire sur les travaux qui se sont déroulés en 1985 à Iasos”, Kazi Sonuçlari Toplantisi 9.2 (1987) (7–11) 8–10; F. Berti, “Travaux à Iasos en 1986”, Kazi Sonuçlari Toplantisi 9.2 (1987) (27–36) 28–29. A detailed final survey report has been produced recently: U. Serin, Early Christian and Byzantine Churches at Iasos in Caria: An Architectural Survey (Monumenti di Antichità Cristiana, Series 2/17) (Vatican City 2004). I pay relatively little attention to this very competent book, as architecture is not my focus and architectural dating via parallels is given very little weight in my study.
Of phasing and dating, a small apsidal ‘martyrion’ was built between tombs established in the agora. On the floor of the martyrion late 4th c. coins were found (6 of Valentinian I, reigned 364–75 and one of Theodosius [?I], reigned 379–95). Above this level a church was built. Both structures incorporated architectural elements taken from the agora: Berti (1986) 28–29. For a detailed discussion of the martyrion see Serin (2004) 32–37 and of the church esp. 38–87. I do not want to comment on the date of the apse of the martyrion (which seals some lamps of apparently later 3rd to early 4th c. date), as this apse is thought to be a secondary construction feature: Serin (2004) 32, with references. The church is dated by Berti on the basis of an unspecified mosaic found in the building that has now disappeared. This mosaic contained an inscription indicating the date [if I have interpreted the report correctly, which is slightly ambiguous here], or which was dated on the basis of the polychrome mosaic style: Berti (1985) 8 (specifying that it covered over the martyrion); Berti (1987) 28 (claiming the date as 5th or early 6th c., specifying that it was part of the church, but that it commemorated the martyrion, noting also the disappearance of the mosaic), whereas Serin (2004) 32 says it was found in the entrance of the north aisle of the basilica. The coins from the martyrion floor underneath the church also provide a TPQ for the church. The study of Serin (2004) 38–87 (discussion), 99–101 (summary), looks for parallels in plan and ornament to date the church to the mid-6th c., which I will not go into here.
Other evidence from around the agora suggests that the plaza was not abandoned before this date and perhaps abandoned a little later. (i) Firstly, inside the southern stoa of the agora, next to the peribolos of Artemis Astias, an encroaching room was built, made out of reused material, including statues and inscriptions. This room overlay a ceramic layer deposited on the stoa paving. This layer contained an African oil lamp and a sherd of LRC, with stamps of little crosses flanking a kantharos, rosettes and lions, which is given as 5th c.: F. Berti, “Mission archéologique Italienne de Iasos: compte rendu des travaux de 1991”, Kazi Sonuçlari Toplantisi 14 (1992) (91–109) 94–95 citing LRP p. 356 fig. 75n–o (lion) and p. 364 fig. 78a (kantharos) [the full range of LRC is 387.5–650 with the neither cited motif here being given a clear date in LRP, although crosses are not generally seen before the 5th c. in such art: see dating foreword]; F. Berti, “The work of the Italian archaeological mission at Iasos in 1998”, in Kazi Sonuçlari Toplantisi 21.2 (2000) (119–26) 166. This suggests that parts of the agora were not transformed until in or after the 5th c. Admittedly, this one room can hardly tell us the full story of the square. Yet, the stoa was probably the first element of the plaza to be occupied by irregular secondary structures, as this was a space that was already provided with walls / a roof. (ii) Secondly, an inscription found outside the east stoa, honouring a praetorian prefect as benefactor of the city. This was dedicated by the boule and demos in the 4th or 5th c. AD (see appendix K9). This suggests the agora was still monumental in the 4th or 5th c. AD (iii) Thirdly, a kiln built against the wall of the church produced unspecified ceramics of ‘Byzantine’ manufacture, believed by Berti to date from the 6th c.: F. Berti “Les fouilles à Iasos en 1989”, Kazi Sonuçlari Toplantisi 12.2 (1990) (235–46) 237–38. Such ‘6th c.’ ceramics can be taken as dating the kiln associatively, using their full date range. This kiln was overlain by a tomb from the subsequent cemetery, which is the only piece of evidence to phase the cemetery, which otherwise is described as producing grave good dating from the 8th c. onwards, and so will not be considered in this study. (iv) Fourthly, in contrast, a hoard of coins of Valentinian II (reigned 375–92) found ‘near the tombs’ around the church does not really help us, as its context is not known, but it does suggest a deterioration of the area at around this time, as such a hoard would not usually be deposited in the middle of an open and functioning civic plaza: C. Laviosa, “La campagne de fouilles de 1984 a Iasos”, Kazi Sonuçlari Toplantisi 7 (1985) (293–97) 294.
Overall, it is the martyrion that really interests us here, not the subsequent church, because it is this building that potentially relates to the disuse of the agora. At present, the only TPQ for the martyrion is the Edict of Milan of 313. I will give it a contextual TAQ of 25 years after 379, derived from the layer found on its floor (so 379–404). Serin (2004) 32 n. 23 expresses reservations about coin dating without specifying what these reservations are, so I will not change the rules of my study here. However, this is a large agora, and the small martyrion might well have existed with continuing civic or market activity elsewhere on the plaza. The best evidence for the decay of the agora comes from the pottery layer covering part of its paving (listed under (i) above), from which we must take a date range of the whole of the 5th c., as ceramics are not specified in more detail. From this can be taken a crude contextual date for the start of disuse of the plaza.
Dating summary (martyrion): range 313–404, midpoint 358.5, class x (historical text), Cs8 (contextual coins), publication 3/3.
Dating summary (start of disuse of agora): range 400–500, midpoint 450, class Cs9 (contextual ceramics), publication 2/3.
Dating summary (appearance of kiln): range 500–600, midpoint 550, class Cs3 (associative, finds), publication 2/3.
13ASI Xanthos (Lycia) (Theatre Square / Roman Agora): A church was built on the Theatre Square. It was surrounded by a cemetery, which both underlies the church and partially overlies its demolished walls. The church is dated to towards the end of the 6th c. AD, on account of the finds in some tombs, which are contemporary with, and in some cases overlie, the ruins of the church [if my reading is correct]. Preliminary reports state that these finds are late 6th to mid-7th c. in date (unspecified ceramics and dress ornaments—earrings, pendentive crosses, bracelets and rings—with an apparent parallel to a hoard from Palmyra of the very end of the 6th to 7th c.): J. des Courtils and D. Laroche, “Xanthos—Le Letoon. Rapport sur la campagne de 1997”, Anatolia Antiqua 6 (1998) (457–77) 463–68, with 465 for Palmyra parallel claim, perhaps referring only to the rings (shown on p. 465 with fig. 8); J. des Courtils and D. Laroche et al., “Xanthos—Le Letoon. Rapport sur la campagne de 1998”, Anatolia Antiqua 7 (1999) (367–99) 372–76; J. des Courtils and D. Laroche et al., “Xanthos—Le Letoon. Rapport sur la campagne de 1999”, Anatolia Antiqua 8 (2000) (399–83) 344 (with mention of accordance of ceramics evidence with the other finds, although a review of the pottery had only just begun). I adopt here the full date range of the finds for the cemetery in dating the church, as at present it does not seem possible to distinguish a more detailed chronology within the late 6th to mid-7th c. I take the finds as indicating an associative date, as they were deposited deliberately as part of the use of the cemetery, not as a levelling fill. On the parallel to the Palmyra hoard, which does not seem to be made in detail anywhere, see A.-M. Manière-Lévêque, “L’évolution des bijoux ‘aristocratiques’ féminins à travers les trésors proto-byzantins d’orfèvrerie”, RA 1 (1997) (79–106) 94, which includes an ‘anneau de lobe’. The hoard is dated by 27 coins dating from Phocas (602–610) to Constans II (642–68). I do not use a regional TAQ of the Persian invasion of 614, as small churches set in former secular public space could well represent a 7th c. development in the context of regional urban patterns.
Dating summary: range 587.5–650, midpoint 618.75, class Cs3 (associative, finds), publication 2/3.
See also (appendix K2b): 13ASI Laodicea ad Lycum (North Agora): A church was built over the open plaza and over the ruins of the north portico: range 494–610, midpoint 552, class Cs2 (catch-all, earthquake), publication 2/3.
V3 Churches Built over Fora / Agorai, the East (Undated)
I have decided not to spend time investigating these in any detail, as there seems to be no rigorous dating basis.
Chersonesus Taurica (Cherson in Crimea): Three or 4 chapels and churches were built over the agora, which have not been dated. The bibliography is reviewed in E. Jastrebowska, “Chersonèse dans l’Antiquité tardive: état des recherches et bibliographie”, AnTard 9 (2001) (399–418) 402.
10MAC Philippi: A church built was over a possible secondary agora, as discerned by geophysics: S. Provost and M. Boyd, “Application de la prospection géophysique à la topographie urbaine, II: Philippe, les quartiers ouest”, BCH 126 (2002) (431–88) 457–73.
13ASI Rhodes: Three churches were built over the agora: E. Kollias, “Η παλαιοχριστιανική και βυζαντινή Ρόδος. Η αντίσταση µιας ελληνιστικής πόλης”, Ρόδος 2400 χρόνια. Η πόλη της Ρόδου από την ίδρυσή της µέχρι την κατάληψη από τους Τούρκους (1523) (Athens 2000) 299–308 (not seen). I thank G. Deligiannakis for this reference.
13ASI Pergamon: A church was built over the agora on the lower slopes of the acropolis: W. Dörpfeld, “Die 1900–1901 in Pergamon gefundenen Bauwerke”, AthMitt (1902) (10–43) 16f and W. Dörpfeld, “Die Arbeiten zu Pergamon 1902–1903, Die Bauwerke”, AthMitt (1904) (116–20) 114ff.
13ASI Knidos (Caria): A church appears to have been built over the agora in R. Özgan, “1997 Knidos Kazısı”, Kazi Sonuçlari Toplantisi 20.2 (1998) (205–209) 208 fig. 1.
13ASI Xanthos (Lycia) (Lower Square): A church was built over the ‘lower square’ (not the same as the Theatre Square): This church has not yet been studied: L. Cavalier, pers. comm.
13ASI Ariassos: A church built over the forum contains a large amount of spolia, including many inscriptions, for which see: S. Mitchell, “Ariassos 1990”, AnatSt (1991) 41 (159–72) 165–66 with fig. 5; H. Rott, Kleinasiatische Denkmäler aus Pisidien, Pamphylien, Kappadokien und Lykien (Leipzig 1908) 25.
13ASI Phaselis: A church built in the tetragonal agora is thought to date from the 5th–6th c. but not after the middle of the 6th c., based on the style of the apse: a half-round projecting apsis rather than a polygonal apsis, common from the 6th c. A date in the 4th c. is excluded, because it is not thought possible to have a church set in a public building at that date: J. Schafter ed., Phaselis. Beitrage zur Topographie und Geschichte der Stadt und ihrer Hafen (Istanbuler Mitteilungen 24) (Tübingen 1981) 96–97. As these arguments are circular, it is better to ascribe the church just as dating sometime within Late Antiquity.
13ASI Seleuceia / Lyrbe: A small chapel was built over a corner of the agora: J. Inan, Toroslar’da bir antik kent: Lyrbe?—Seleukeia? = Eine antike Stadt im Taurusgebirge. Arkeoloji ve Sanat Yayinları (Istanbul 1998) 36–37, 112 fig. 5.
15ORI Kelenderis: A church was built over the agora: K.L. Zoroğlu et al., “Kelenderis 2002 yılı kazısı raporu”, Kazi Sonuçlari Toplantisi 25.2 (2003) (451–66) 455–57.
Discounted: 10MAC Dion: A basilical church at Dion is described by Hoepfner as being built on the site of the agora “auf dem Gelände der Agora”. However, it is actually located behind a portico of the square. It was not built within the plaza, as the square continued to be used within the Late Roman period: W. Hoepfner, “Das Ende der Agora”, in Die spätantike Stadt und ihre Christianisierung. Symposion vom 14. bis 16. Februar 2000 in Halle/Saale, edd. G. Brands and H.-G. Severin (Spätantike—Frühes Christentum—Byzanz. Kunst im ersten Jahrtausend. Reihe B: Studien und Perspektiven 11) (Wiesbaden 2003) (145–50) 148. There may be more recent studies of its date, but it is not pertinent here.
See also notices of overbuilding / transformation of fora in Africa by fortifications in the main text: for the reconquest period in the west (Thugga (Dougga), Madaurus, Tubernuc, Sbeitla, Abthungi, Mustis, Sua and Lepcis Magna) and for the same period in the east (Epidauros, Aizanoi and Miletus).
V4a Degradation and Disuse of Fora / Agorai in the West, Late 3rd-4th C.
01BRI Gloucester: The latest courtyard surface was a patchy resurfacing in broken tile and crushed limestone. It can be dated from ca. 250 onwards, because this latest surface has set within it a phase of repair to a statue base undertaken in reused material, which implies a nominal TPQ of the mid-3rd c.: see appendix K4d. Above this latest surface was a black gritty loam containing a small amount of building rubble, some sherds of (unspecified) 4th c. pottery, two coins, and a lot of bones. The lack of building rubble and the fact that this deposit covered areas where stone slabs had been removed suggested to the excavator that the forum had been dismantled in an orderly fashion prior to the end of occupation on the site [the black loam seems to represent abandonment]. Three squared posts were driven into the courtyard make-up layer, and there was a 40 cm thick accumulation of earth over the plaza, both before the Late Saxon occupation of the same area: H. Hurst, “Excavations at Gloucester 1968–1971: first interim report”, AntJ 52 (1972) (24–69) 58. We do not know whether the 4th c. pottery was in small fragments (and so perhaps deposited in the 5th c.), or if it occurred in large pieces (more likely discarded in the 4th c.). The presence of several fragments, with nothing else earlier or later, gives us a date via associative finds of 300–400, the full range of the unspecified pottery, although the dearth of 5th c. material culture in sub-Roman Britain makes it difficult to be sure.
Dating summary (for beginning of abandonment): range 300–400, midpoint 350, class Cs3 (associative, finds), publication 3/3. Poor.
01BRI Silchester: Industrial activities were carried out in the civil basilica in the period ca. 259–84, based on associative coin finds: see appendix X1b. Contemporary occupation of the forum plaza surface, as revealed by an excavation trench only a few metres wide, is less dramatic. It consists of secondary surfaces and charcoal spreads, although these do suggest degradation in the forum as traditionally conceived. The latest surface produced a coin of Eugenius (of AD 392–95) and was cut by a number of stake holes and small postholes: M. Fulford and J. Timby et al., Late Iron Age and Roman Silchester: Excavations on the Site of the Forum Basilica 1977, 1980–86 (Britannia Monograph Series 15) (London 2000) 75 with fig. 71. A pit was cut into the foundations of the colonnade, from which a coin of 360–68 was recovered (pp. 76–78). However, as a single find it only provides a TPQ for the decay of the structure, not any contextual dating.
Dating summary (for beginning of degradation of forum): range 259–84, midpoint 271.5, class Cs3 (associative, phase of development), Cs3 (associative, coins from basilica), publication 3/3. Poor.
02HIS Baelo Claudia (near Cádiz): The civil basilica collapsed and the macellum was turned into a rubbish tip, with a settlement established on a different street alignment. This seems to be the result of an earthquake, as deduced from the collapse pattern of the basilica, deformations of forum slabs and other features in the area, with ceramics and coin evidence pointing to dislocation sometime in the late 3rd to late 4th c., perhaps most likely in the second half of the 4th c.: P. Sillières, Baelo Claudia, una ciudad romana de la Bética (Madrid 1997) 57–58; 61–63; P.G. Silva et al., “Archaeoseismic record at the ancient Roman city of Baelo Claudia (Cádiz, south Spain)”, Techtonophysics 408 (1–4) (2005) 129–46 (who record 350–95 as the most likely period for the quake, based on ceramics from the city walls and macellum). N.B. Neither of these reports contains ceramic or coin evidence. The last references for the city wall and macellum are: F. Didierjean et al., “La douzième campagne de fouilles de la Casa de Velazquez à Belo en 1977 (Bolonia, Cadix)”, Chro. Mél. Casa de Velazquez 14 (1978) 433–464 (not seen); L. Menanteau, Belo II : Belo et son environment (Detroit de Gibraltar), Etude physique d’un site antique (Casa de Velazquez, Serie Archeologie 4) (Paris 1983) (not seen). I take the full range given for ceramics and coin finds as associative finds, as given above, to indicating the dating range of this destruction of the forum.
Dating summary: range 287.5–387.5, midpoint 337.5, class Cs3 (associative, coins from basilica), publication 2/3.
02HIS Astigi (Écija, Andalucia): Buildings around the forum were progressively substituted by housing, but the main plaza remained open and intact: S. Ordoñez, Colonia Augusta Firma Astigi (Écija 1988); P. Saez et al., Carta arqueológica municipal. Écijia, 1: La cuidad (Seville 2004) 34 (not seen); A. Romo et al., “De las termas a la mackbara. Intervención Arqueológica en la Plaza de España de Ecija (Sevilla)”, Anuario Arqueológico de Andalucía (1998.3) (979–96) 979 (not seen).
Dating summary: Pending.
02HIS Aurgi (Jaén, Andalucia): In an area thought to be the forum (not confirmed by architectural finds), a paved area was resurfaced with orange earth / mortar 5–12 cm thick, before being converted into a rubbish dump, indicating the decay of the plaza: J.-P. Bellón and C. Rueda, “De foro a vertedero? Perdidos en el Decumanus Maximus de Aurgi. Resultados de la intervención arqueológica de urgencia en la calle Santo Domingo, 19 a 25, de Jaén”, Arqueología y Territorio Medieval 8 (2001) (175–86) 179.
The dating of the orange surface depends on a layer under it, which produced ceramics given as dating to the first half of the 4th c. (TS Tardía Meridional, imitation of Form 54 Lamboglia [Lamboglia 54 = Hayes 61(A) in LRP gives 325–400/420 for Hayes 61A with SLRP giving the range 380/390–450 for type Hayes 61b, changed to 400–500 by Bon]; ARS C and D; paleochristian painted ware). I am not able to comment on the dates of local imitation wares, so must accept the range given and derive a rough contextual date for the orange floor surface, using the whole period of the ceramics cited.
The dating of the subsequent rubbish dump, which included dark earth and carbonised organic material, depends on contextual ceramics. It produced ceramics believed to be of the second half of the 4th c. and beginning of the 5th c.: ARS C and D; TS Tardía Meridional [likely Terra Sigillata Hispánica Tardía Meridional, which I was not able to check the date of]; ‘paleochristian’ [ditto]; Lamboglia 40 [Lamboglia 40 = Hayes 50A and dated as 300–60 in Atlante p. 65, quoting Hayes; then LRP pp. 71–73 gives 230–360 for the range of two forms of Hayes 50A. However, Hayes 50 has recently been redated to 350–400 (E. Vaccaro pers. comm., which I defer to]. It is difficult to identify the early 5th c. ceramics in this group from the detail provided, meaning that we have to accept the full date range given for this group, as the last finds, in order to date the layer. The nature of the layers led the excavators to assume that this deposit was a primary rubbish deposit, laid over the plaza. As a primary rubbish deposit, this layer could have begun by being deposited anytime within the range of the ceramics, rather than all together at a time after the last ceramic was added. Thus, despite a few wares of which I cannot confirm the date, I will defer to the date of the ceramic range provided by the excavator and use this full range to suggest a weak contextual date for the rubbish deposition of 350–412.5. This range does not contradict the ranges of the wares for which I have been able to confirm the dates.
Dating summary (final orange surface): range 300–50, midpoint 325, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (rubbish deposition): range 350–412.5, midpoint 381.25, class Cs9 (contextual ceramics), publication 2/3 (as contextual control within the report is not strong).
02HIS Emerita (Colonial Forum): In the south-west corner of the Colonial Forum, brown earth surface UE 210 was established in the inner open space of the plaza (los espacios forenses interiores). This had an abundant ceramic material datable from the second half of the 4th c. and the whole of the 5th c., but I was not able to see diagnostic wares listed in the ceramics report as the report stated, for this context. This surface predated fill layers in robbing trenches of an adjacent forum building (UE 164, UE 191, UE 193, UE 343, UE 344, UE 347, and UE 348)) which contained identical forms of TSH Tardía and ARS D) which led the excavators to assume a rapid spoliation in the 5th c. AD: R. Ayerbe Vélez, “Solar de la calle Juan Dávalos Altamirano, 6–10 (esquina calle Los Maestros) ” in El foro de Augusta Emerita: genesis y evolución de sus recintos monumentales, edd. R. Ayerbe Vélez, T. Barrientos Vera, and F. Palma García (Merida 2009) (142–73) 153–54.
Details of the ceramics are available in X. Aquilué Abadías and J. Ramón Bello Rodrigo, “Los materiales arqueológicos de época romana y tardorromana procedentes de las excavaciones del foro colonial de Augusta Emerita”, in the same volume (405–444) 409. Here it is clear that only the following fill contexts from the robbing trenches have significant ceramic finds: UE 164, UE 193, and UE 343. UE 164 contains only a fragment of TS Sudgálica form Drag. 27, of 1st c. date. UE 193 contained the following fine wares: a few fragments of ARS D, forms Hayes 61A [ca. 325–380 in LRP with SLRP], Hayes 91A/B [not in my list], along with TSTH form Drag. 37 tardía [this could be forms TSTH Paz 4.23–4.26 of TSHT Dragendorff 37 of a date range of 250–380 in fig. 14 table 2 of Paz]. UE 343 contained finewares of Hayes 104B [540–600 Bon] of ARS D2 and amphoras of form Keay 23 [200–?450 RADR], ánforas African amphoras of Keay 25H [I can’t find this, neither Keay nor in Bonifay] and eastern amphoras of Keay 54 bis [not in Bonifay].
More spoliation and rubbish contexts are mentioned on pp. 409–410. These include UE 464, UE 484, and UE 494, all rubbish layers from inside the plaza. UE 464 contains a fragment of a stamped base of ARS D [300–700 DCAMNO], with a lattice of style A-2 or 3 of Hayes, comparable to motif Atlante 31=Hayes 69 [ca. 425–450 in LRP]. UE 484 contains rims of forms 1st–2nd c. forms. UE 494 contains fragments of ARS C [200–500 DCAMNO] and ARS D [300–700 DCAMNO].
A general dating commentary for the forum based on these ceramics is given on p. 411, with further ceramics that apparently give a date for the abandonment of this part of the colonial forum in the first half of the 5th c. Unfortunately, it is not clear to me where many of these ceramics come from, so I will only use the stratified finds above. Furthermore, I am not able to see how the fill contexts of robing trenches listed earlier support a date in the 5th c. for spoliation. UE 193 seems to contain ceramics with a range within 250–380, although I could not obtain a clear date for Hayes 91A/B, a ceramic form not listed in Hayes LRP or Hayes SLRP or in Bonifay. This was a long and difficult report, of difficult small urban trenches, without an easy summary of phasing that would have made it easy to make the connection between key contexts and finds. Thus, I must admit defeat until I can obtain help to understand the report better.
Another part of the colonial forum was excavated by P. Mateos Cruz and A. Pizzo, “Intervención arqueológica en el solar ocupado por los restos del templo de la C / Holguín”, in El foro provincial de “Augusta Emérita”: un conjunto monumental de culto imperial, ed. P. Mateos Cruz (2006) (122–138) who notes on p. 205 that the (largely intact) pavement of the colonial forum was covered progressively with rubbish dumps (UE 47 in area 1 and UE 43 in area 2). However, the layers are simply described as ‘Late Roman’, with mention in UE 43 of ARS and in UE 47 of three ARS fragments and one piece from a lamp, on pp. 202 and 200, respectively. This level of information of course provides no basis for substantiating the Late Roman dating.
Dating summary (decay of forum): Pending, likely prior to 380, class Cs9 (contextual ceramics), publication 3/3.
02HIS Emerita (Forum Adiectum / Eastern Platform, next to Colonial Forum): There are two late antique phases.
In the first late antique phase, the excavators have identified part of a large well-constructed building, of granite ashlars, with buttresses, of which three rooms have been uncovered, employing many reused marbles taken from the forum in the southern part of the Eastern platform area, which has been given a date in the mid-5th c., from finds associated with its construction. However, details of this building have not been fully published yet: R. Ayerbe Vélez, T. Barrientos Vera and F. Palma García, “Los complejos forenses de Augusta Emerita” in El foro de Augusta Emerita: genesis y evolución de sus recintos monumentales ed. R. Ayerbe Vélez, T. Barrientos Vera, and F. Palma García (Merida 2009) (667–831) 829–30. The building’s walls are described in detail as structure A18, in R. Ayerbe Vélez, “Solar de Trav. Parejos-Trav. Hernán Cortés”, in El foro de Augusta Emerita: genesis y evolución de sus recintos monumentales, edd. R. Ayerbe Vélez, T. Barrientos Vera, and F. Palma García (Merida 2009) (249–94) 271–75 with 272 fig. 204 giving a plan of the building. An earth road built past the building on its east side is described on p. 271 as having a granular surface, very compact and hard, with a concave surface, ca. 5 m wide, although it looks wider on p. 272, fig. 204, closer to 6 m, which might include unnavigable edges of the roadway. The ceramic report is X. Aquilué Abadías and J. Ramón Bello Rodrigo, “Los materiales arqueológicos de época romana y tardorromana procedentes de las excavaciones del foro colonial de Augusta Emerita”, in the same volume (405–444) 433.
Of dating for the abandonment of the forum, the ceramic report makes clear that the late antique building was constructed above the abandoned Early Imperial structures of the Eastern Platform. Abandonment contexts with finds listed are UE 667 and UE 669. UE 667 contains a fragment of ARS C [200–500 DCAMNO] and a base of Drag. 37 of TSTH [250–380 in Paz] and other pottery including Late Roman cooking ware. UE 669 contained ollas of coarse reduced pottery of Late Roman date. This evidence could be used to produce a weak contextual ceramic date of 250–75 for the abandonment based on the start date of the latest datable sherd, but given that this is the only closely datable sherd from the context, this was not very wise, especially as it has a number of sub forms in Paz. Thus, a date range of 250–380 for the abandonment of the forum is perhaps better, using the full range of the ceramic form to produce a rough contextual date, as is done at Sagalassos for example with phases of SRSW.
Of dating for the well-constructed building, the ceramic report notes that construction contexts for the subsequent building contain some residual pottery, which I will not list here. However, UE 656, a construction level of the earth street on the exterior of the granite building, is very rich in late material. UE 656 contains a large number of ceramics including the following diagnostic wares: (i) ARS C-2, forms Hayes 50 [350–400 E. Vaccaro pers. comm.]; (ii) ARS C-3, forms Hayes 53 [350–430+ in LRP for 53A which has the longest range] and Hayes 73 [387.5–475 in LRP]; (iii) ARS D forms Hayes 61A [ca. 325–380 in LRP with SLRP], Hayes 67 [350–500 in LRP Bon], Hayes 76 [ca. 425–75 in LRP] and Hayes 91A [400–450 Bon] with bases stamped of style A-2 of Hayes centric circle motifs of type Hayes 28=Atlante 12 alternating with palmetes of types Hayes 1–3=Atlante 108–112 and style A-2/3 with lattice square motifs of type Hayes 69=Atlante 31; (iv) Lamps of ARS D form Atlante VIII A [350–450 Bon for Atlante VIII A1c / A2b with widest date range of subtypes] with a decorative chrism comparable to those on lamps of Dressel 30 [Dressel 30 lamp is 290–410 on the British Museum website, without references]; (v) TSTH form Drag. 37 tardía [250–380 in Paz] and Hispánicas Lisas of forms 4, 6, 10 and 66 [I don’t know what date this type of pottery is]; (vi) Beatican Amphoras of Keay 13 A=Dressel 23 [200–512.5 RADR] and South Spanish amphoras of Keay 16=Almagro 50 [187.5–450 ICAC]. Taken together, this evidence suggests that the road next to the building was laid down in 425–50, reflecting a contextual date based on the start date of the latest ceramic, which is ARS D Hayes 76 of ca. 425–75 in LRP. See similar dating arrived at for this context in the ceramic report on pp. 441–442 by studying other ceramic reports, outside the manuals used in this study. I was not able to check the position of UE 656 in the matrices published for the site, partly because complex stratigraphy is subsumed into named features, without feature matrices being provided, or perhaps I am just a bit slow-witted and did not see them. However, the report seems to suggest that the ‘Late Roman road’ is built respecting the new building, and so is later than it.
Overall, the dating of the well-constructed building can be given a date after 250, the earliest date for the abandonment of the forum, but before the construction of the road, not later than 450, based on the upper range of the contextual date that I have ascribed to it.
In the second late antique phase, dwellings were built here and rubbish dumps were established according to P. Diarte-Blasco, La configuración urbana de la Hispania tardoantigua. Transformaciones y pervivencias de los espacios públicos romanos (s. III–VI d. C.) (BAR International Series 2429) (Oxford 2012) 54 notes that, citing R. Ayerbe Vélez, “Solar de Trav. Parejos-Trav. Hernán Cortés”, in El foro de Augusta Emerita: genesis y evolución de sus recintos monumentales, edd. R. Ayerbe Vélez, T. Barrientos Vera, and F. Palma García (Merida 2009) (249–94) 265–67, which on p. 267 points to a large quantity of late antique ceramics in dumps associated with a wall UE 342 of domestic structure (A19) established within a passage. Dump UE 493 is the first dump of these dumps (UE 408, UE 487, UE 490 and UE 493). Disappointingly, ceramics from UE 493 are described in the specialist report only as being unspecified coarse wares, accompanied by fragments of marble and tiles: Aquilué Abadías and Bello Rodrigo (2009) 409.
Dating summary (abandonment of the forum): range 250–380, midpoint 315, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (well-constructed building): range 250–450, midpoint 350, class Cs3 (associative, phase of development), Cs9 (contextual ceramics), publication 3/3.
Dating summary (roadway past building): range 425– 50, midpoint 437.5, class Cs9 (contextual ceramics), publication 3/3.
02HIS Emerita (Provincial Forum) There was apparently a spoliation in the 4th to 5th c., followed by buildings set over the plaza in the 5th c. (decay dates to between Maximian and domestic occupation): P. Mateos, “Excavaciones en el solar de la C/Almendralejo y Holguin”, in El ‘Foro Provincial’ de Augusta Emerita: un conjunto monumental de culto imperial, ed. P. Mateos (Madrid 2006) (68–110) 114 117 (not seen). In the absence of this report I am not able to comment further.
Dating summary: Pending.
02HIS Segobriga: The bronze letters of a great inscription set in the paving were robbed and a layer accumulated on the forum paving: J.M. Noguera Celdrán, J.M. Abascal Palazón, and R. Cebrián Fernández, “Programa escultórico del foro de Segóbriga”, in Escultura romana en Hispania, vol. 5 (Actas V Reunión Escultura Romana en Hispania, 9 al 11 de noviembre de 2005), edd. J.M. Noguera Celdrán and E. Conde Guerri (Murcia 2008) (283–343) 300. Into this layer a togate statue fell from its base. In the report the layer is dated to between the second half of the 3rd c. and the end of the 4th c., based on the finding of terra sigillata clara C (i.e. ARS C) [200–500 DCAMNO] and Hayes 50: p. 300 with n. 94 and fig. 12. Elsewhere the authors of the report talk about the spoliation of the basilica in the 4th c. (p. 322) and put the collapse of the basilica between the second half of the 4th c. and the early 5th c. (p. 302), noting a dump in the basilica of 5th c. date, containing a statue fragment, although in none of these latter three cases dating evidence is provided. The dating of Hayes 50A in LRP/SLRP was 230–350, but 50B is now dated to 350–400, whereas 50A/B dates to 300–60 according to E. Vaccaro pers. comm. 2016 [Hayes 50B is 350–400+ in LRP p. 73]. Given that we do not know what forms of ARS C we are dealing with, we have to accept the full dating range given in the report, based on unspecified ceramics for the decay of the forum, as associative finds, as the earth layer likely built up slowly rather than being deposited as a dumped fill at one moment, based on associative finds.
Dating summary: range 250–400, midpoint 325, class Cs3 (associative, finds), publication 2/3.
06ITS Grumentum: A sequence of denudation, repair and demolition has been detected in one part of this forum: A. Mastrocinque et al., “Gli scavi dell’Università di Verona nel foro di Grumentum (Potenza). Anni 2007–2009”, Fasti Online (2010) 1–29: http://www.fastionline.org/docs/FOLDER-it-2010-183.pdf p. 16 (for the robbing layers); A. Mastrocinque, “Indagini nell’area dell foro di Grumentum”, Fasti Online (2006) 1–7: http://www.fastionline.org/docs/FOLDER-it-2006-69.pdf p. 4 (for the foundry). The dating of these developments is difficult, as the reports of recent excavations are only available in preliminary form. Robbing layers have been found in the north-east corner, suggesting some degradation. A foundry was established in a room off the south portico, but was not the last phase of forum buildings. The foundry is dated by unspecified ceramics in the foundation layer, which extend from the Republic up to the 3rd–4th c. AD but not beyond, which I will take as providing a date of the whole of the 4th c., which is what the 2006 report on p. 4 asserts. The phase after it (raised floors in rooms and a new portico floor) is not dated, although the 2006 report p. 4 suggests it is probably after the end of the 4th c. Levelling and robbing layers on the forum are dated only from the 4th to the first half of the 6th c., based on unspecified ceramics: 2010 report p. 16. Whilst the foundry provides important evidence for a phase of occupation, it represents a temporary degradation, rather than decay of the plaza. Rather, the layers associated with spoliation provide a stronger indication for the beginning of decay. This must have happened sometime after 300 (based on the foundry), but prior to the middle of the 6th c. (based on ceramics from the robbing layer). Without more detail I am obliged to give the full range given for the ceramics here as dating.
A summary of the excavations east of the (?east) portico of the forum in 2010 suggested a collapse and robbing of walls in the 5th–6th c. However, the evidential basis for this interpretation (whether ceramics or other finds?) is not yet clear, so I will not use it here: A. Mastrocinque, http://www.fastionline.org/excavation/micro_view.php?fst_cd=AIAC_468&curcol=sea_cd-AIAC_3321 (last accessed Jan 2016).
Dating summary (foundry): range 300–400, midpoint 350, class Cs9 (contextual ceramics), publication 2/3.
Dating summary (raised floors in rooms and portico repaving): range 300–400, midpoint 350, class Cs9 (contextual ceramics), Cs3 (associative, phase of development), publication 2/3.
Dating summary (spoliation): range 300–550, midpoint 425, class Cs9 (contextual ceramics), publication 2/3.
06ITS Scolacium: After a hypothetical earthquake in the 4th c. the forum was occupied by precarious buildings in wood, alongside fences, indicating the decay of the plaza. These features are indicated by postholes cut into the paving. In the phases leading up to the abandonment of the plaza the forum lost its public function and became an artisanal area: R. Spadea, Scolacium: una città romana in Calabria: il museo e il parco archeologico (Soprintendenza per i beni archeologici della Calabria 2005) 53 and 70. No chronological indicators are available in this text, and I was not able to locate any reports relating to these discoveries. However, we seem only to have a belief that the occupation of the forum extends into the 5th c., as a coin of Honorius of 408–23 was found on the paving, and that the ‘latest’ building activity should thus be placed in the 4th c.: C. Donzelli, “Le strutture tardoantiche di Scolacium”, MEFRM 103.2 (1991) (485–503) 496–501. Therefore, I place the artisanal and abandonment features in the part of the 5th c. after the start date of the last published find. It is possible that the forum became an artisanal area prior to abandonment, but the exclusion of such details from the report dealing with later 3rd to 4th c. features suggests that it belongs in the period when the forum was experiencing decay.
Dating summary (artisanal and decay phase): range 408–500, midpoint 454, class Cs7 (TPQ, coin), Cs3 (associative, phase of development), publication 2/3.
See also (appendix K2a): 03GAL Amiens: In the eastern courtyard of the forum dumps of slag and a crucible containing copper or bronze indicate decay, in favour of an industrial use of the forum in 364–89, prior to abandonment in the period 410–35 (based on contextual coins and contextual ceramics).
06ITS Ostia (Palaestra of the Forum Baths): In phase 7 rooms were built inside the portico, at a higher occupation level than the ancient plaza. This can be treated as ‘decay’, although the revetment had already been stripped off in an earlier phase. This probably happened in the period 443–?450.
Discounted: 01BRI London (Londinium): The supposed demolition of the forum, ca. AD 300, will not be examined here, as the areas excavated are small and the evidence is somewhat contradictory, making a detailed analysis of little value; 03GAL Bavai was given up in around 260/70 and a fortification was established shortly after.
Elsewhere in Hispania, Gaul and Italy: Many cities such as Cambodunum, Augst and Glanum with large fora show little signs of occupation after the 3rd c. See sites that are listed in Diarte-Blasco P. (2012) La configuración urbana de la Hispania tardoantigua. Transformaciones y pervivencias de los espacios públicos romanos (s. III–VI d. C.) (BAR International Series 2429) (Oxford 2012) (e.g. Ercavica or Pollentia) and also A. Ferdière ed., Capitales Éphémères. Des capitales de cités perdent leur statut dans l’Antiquité Tardive (Revue Archéologique du Centre) (Tours 2004). I am not looking at them here. 02HIS Celti (Hispania): A forum, identified based on the plan form, was demolished to build a private house in the late 2nd c. / early 3rd c. For this development dating includes a sherd of TSCLC1 pottery (given as beginning production around AD 200) from inside a mud-brick column, and a bronze coin of Commodus of AD 190/91 that was squared, suggesting it had been in use for a long time: S. Keay with S. Gibson and K. Wilson, “Reconstruction of the site”, in Celti (Peñaflor). The Archaeology of a Hispano-Roman Town in Baetica. Survey and Excavations 1987–1992, edd. S. Keay, J. Creighton and J. Remesal Rodríguez (Oxford 2000) 85–86, 176–92; 06ITS Cosa and 06ITS Monte Iato both seem to have become very different in character by the end of the 3rd c., resembling villages rather than cities in their level of occupation, so are best not covered here.
Discounted: 02HIS Corduba (Colonial Forum): P. Diarte-Blasco, La configuración urbana de la Hispania tardoantigua. Transformaciones y pervivencias de los espacios públicos romanos (s. III–VI d. C.) (BAR International Series 2429) (Oxford 2012) 132, notes that the colonial forum has honorific dedications down to the time of Constantius II (CIL 2.2/7 265) but that the paving is covered with waste at end of 4th c. I cannot see dating evidence for this in the report that she cites: I. Carrasco Gómez, “Intervención arqueológica de urgencia en un solar sito en calle Góngora número 13 esquina a calle Teniente Braulio Laportilla (Córduba)”, Annuario Arqueologico de Andalucía 3 (1997) (199–208) 207 (which notes disuse and dumps (vertidos) ‘probably already in the 4th c. ’, without citing supporting evidence. 02HIS Corduba (Forum Adiectum): P. Diarte-Blasco, La configuración urbana de la Hispania tardoantigua. Transformaciones y pervivencias de los espacios públicos romanos (s. III–VI d. C.) (BAR International Series 2429) (Oxford 2012) 132 notes that paving is dismantled in the 4th c. citing R. Garcia and I. Carrasco, “Hallazgos en el número 5 de la calle Morería y nuevo espacio público en Colonia Patricia”, Anales de Arqueología Cordobesa 15 (2004) (145–72) 169, which notes that this process of degradation, along with the placing here of dumps (vertidos), starts at the end of the 3rd or beginning of the 4th c., without providing any evidence to support this dating.
V4b Degradation and Disuse of Fora / Agorai in the West, 5th C.
01BRI Exeter: The forum was demolished and then abandoned, as a cemetery was laid out over it: see P.T. Bidwell, The Legionary Bath-house and Basilica and Forum at Exeter (Exeter 1979) 110–13. A TPQ for the demolition comes from the latest repair to the basilica, when in period 3A a mortar floor was laid down, containing a coin of Valens (of AD 365–78). Further dating comes from a follis of Diocletian (reigned 284–305) from the robbing trench of the tribunal screen. A good TAQ for the demolition comes from two C14 dates for burials in the cemetery, centred on the 5th c., whilst a third date in the 11th c. is disregarded as anomalous. These two burials were cut into the basilica floor and have yielded 5th c. dates of AD 420 +/-70 (HAR-1614 1530 +/-70 BP) and AD 490+/-80 (HAR-1613 1460 +/–80 BP): Bidwell (1979) 104–111. I will use a nominal midpoint of 455 to date the burials, to simplify matters, thus providing a TAQ for the demolition of the forum. I include this example because, unlike other cities in Britain where fora are abandoned, here there is a deliberate demolition followed by some occupation, rather than just a general abandonment of the city.
Dating summary (abandonment of forum): range 365–455, midpoint 410, class Cs7 (TPQ coin), Cs10 (scientific, radiocarbon date), publication 3/3.
02HIS Regina (Extremadura): Industrial and domestic occupation continued in the forum until the beginning of the 5th c., when the drains were blocked and earth was built up, indicating decay: J.-M. Alvárez and J.L. Mosquera, “Excavaciones en Regina (1986–1990)”, Extremadura Arqueológica II (Badajoz 1991) (361–71) 362–64 and 368 (not seen).
Dating summary: Pending.
02HIS Tarragona (Tarraco) (Provincial Forum): This large plaza has been described as in decay, invaded by 5th c. rubbish dumps and houses. However, much of this evidence comes from the adjacent temple platform, or from outside the forum. Of the square itself, relatively little has been studied. Excavations have taken place at one small spot at the centre of the eastern side of the plaza and in two staircase towers (the Antiga Audiència on the south-west corner and the ‘Pretori’ / Torre di Pilato on the south-east corner).
At the centre of the eastern side of the plaza, an excavation produced architectural pieces, thought to relate to the dismantling of the forum, but did not produce significant ceramics [apparently, as they are not mentioned]: Taller Escola d’Arqueologia de Tarragona, “El fòrum provincial. Estat de la qüestió”, in Un abocador del segle V d. c. en el fòrum provincial de Tarraco (Tarragona 1989) (435–88) 446–48, esp. 487, with resumé in English on 457–59. A dump (UE 918), containing bones, glass, coins and lots of ceramics, came from close by. It was recovered from the excavation (p. 80) on the crossing of Carrer de les Cuirateries and Carrer d’en Vilarroma [that seems to have occupied the area of Plaça de Mossèn Salvador Ramon i Vinyes and the Lamorada restaurant: I checked p. 81 fig. 34 against Google Streetview]. This area corresponds with a part of the forum paving in the south-east corner of the complex, in a plot that reached the edge of the portico pilasters, as noted on fig. 34. It was laid down on the site of the robbed pavement of the forum, at a level lower than the height of the original paving, as the robbing had descended 2 m below the forum: p. 85.
The dump was dated to the years 440–50 and reflects a single moment of deposition rather than a deposit that accumulated over a long period: pp. 423–33. This dating (summary p. 425) was based especially on the ARS and amphora characteristics of the first half of the 5th c., alongside a minority of forms that began around AD 440 and became more abundant and significant as one advances in the second half of the 5th c. Especially important to the excavators are ARS D Hayes 87A [450–500 in LRP], 87B [500–512.5 in LRP] and 99 (noted as of controversial chronology, but starting to be documented in the second quarter of the 5th c.) [487.5–612.5 in Bon with variants of shorter range], alongside certain African amphora of ‘Vandalic’ type Keay 61 [587.5–650 Bon for all variants: there may be an error here] and 63 [which I was not able to find a date range for, although Keay 1984 had one example deposited in the middle or later 6th c.], which according to the authors appear with the occupation of North Africa by the Vandals in 439. These are the only wares that suggest a date very advanced in the first half of the 5th c. In contrast, the absence of certain forms and certain decorative motifs, alongside the absence of Late Roman C, prevent the report writers from setting the context in the second half of the 5th c. The lamps and coins suggest a moment in the first quarter of the 5th c. However, the lamps do not have the classic African form Atlante X [Bon 425–700, with many subtypes], which are omnipresent in all stratigraphic contexts after ca. 425, or its imitations. Furthermore, the coins seem to reflect the 4th c., with an absence of AE4, characteristic of the 5th c. The report writers note the [non-residual] presence of ‘T.S. Lucente’, which is found as a residual ceramic in the second half of the 5th c. The finds are described in more detail on pp. 425–29. I feel in this case that deference is warranted, given strong local knowledge of the ceramics, so I accept the contextual date for the dump that is offered in the report, to the years 440–50, even though the presence of Hayes 99 would normally push me to propose a date of 487.5–512.5, as this is a late ware recently redated by Bon, a re-dating that the excavators were not aware of in their report of 1989.
The Antiga Audiència was spoliated (in phase IV of the excavation sequence) for decorative and structural elements in the second half of the 5th c., followed by the accumulation of rubbish deposits: X. Dupré i Raventós and J.M. Carreté i Nadal et al., La ‘Antiga Audiència’. Un acceso al foro provincial de Tarraco (Madrid 1993) 82. These rubbish deposits are dated based on finewares and amphoras to the second half, and most probably final quarter, of the 5th c. The pottery is detailed in X. Aquilue Abadias, “Las ceramicas finas de los niveles tardo-romanos”, in the same volume 117–50, and J.A. Remolà i Vallverdú, “Los anforas de los niveles tardo-romanos”, in the same volume 151–72 (not seen). I will only summarise the ceramics here from Aquilue Abadias (1993), who describes the rich finds of the rubbish deposits in detail. He notes (p. 144) that forms of ARS D1 and D2 place the deposits in the second half of the 5th c., most probably the last quarter, as demonstrated by the presence of Hayes 79 [400–500 in LRP, possibly 350–500 in Bon], Atlante 40.9 [500–512.5 Atlante], Hayes 86.1= Lamboglia 60bis [Hayes 86 is 487.5–512.5 in LRP], Hayes 80B/99 [500–512.5 Atlante], Hayes 12.1 [Hayes 12 is reclassified as 12/102 and 12/110, of which this is presumably the former, so now 450–512.5 SLRP], Hayes 12/110 [450–550 Atlante] and Atlante 53.2 [Atlante = Hayes 104C according to Atlante, which is 550–650 in Bon], which are not encountered in well-dated contexts of the first half of the 5th c. The absence of plates of Hayes 104, said to be abundant in the early 6th c. contexts [Hayes 104 is 487.5–650 in Bon, but with many variants of shorter date ranges], suggests that these deposits date before this time. There are some forms characteristic of the first half of the 5th c., like Hayes 67 [ca. 360–470 in LRP], Hayes 91 A/B [Hayes 91 is 350–650 based on LRP, not contradicted by Atlante or Bon, who lists the many variants with shorter date ranges] and Hayes 80/81 [both 80 and 81 are 440–500 in LRP], but these are in a minority. There are also plates similar to the form Hayes 84 (Antiga Audiència 86/87) with feather rouletting [84 is ca. 440–500 in LRP], which is characteristic of the 5th c. and not easily posterior to AD 500. Also, the presence of Hayes 3 [60–150 in LRP] with variants Hayes 3C [100–150 in LRP, but possibly 287.5–?250 Atlante?] and Hayes 3E [I was not able to check this form], suggests a dating towards the second half of the 5th c., corroborated by the stamped motifs dated to the period 440–90. I will defer to the report’s dating for the overall formation of this deposit, of 475–500, as this is a sophisticated report, reflecting knowledge of regional stratigraphy, and the re-datings, which could put the date slightly later when they went to press. Unfortunately, I could not find details of the ceramics in the spoliation deposits (as distinct from the rubbish deposits), although the finds of these layers are mentioned as placing the spoliation process in the second half of the 5th c., according to Dupré i Raventós and Carreté i Nadal (1993) 82 (referring to appendices 4.2 and 4.3 in the same volume).
The Torre de l’Audiència site was apparently set within the same stair tower. Excavations here produced other deposits of the second half of the 6th c., sealed under an opus signinum floor: S.J. Keay, Late Roman Amphorae in the Western Mediterranean. A Typology and Economic Study: The Catalan Evidence (BAR-IS 196) (Oxford 1984) 17–19, with figs. 9.1 and 9.2 displaying the types of ARS present. I will not detail the ceramics here, as a positive change of function of a staircase tower does not amount to a clear decay of the forum, in the same way as does the conversion of the stair tower into a rubbish dump, as recorded in the previous site. Thus, it is not critical for the argument studied in this appendix. However, I can say that the deposit includes ARS Hayes 106 [which in LRP/SLRP is ca. 600–660+], alongside other similarly late wares.
The ‘Pretori’ / Torre di Pilato site revealed abandonment layers, of which the first to have finds had unspecified bronze coins of the end of the 4th c. There was also a large quantity of unspecified ceramics, and late terra sigilata is mentioned, see: A. Balil, Excavaciones en la “Torre de Pilatos” (Tarragona). Campañas de excavaciones de 1962 (EAE 65) (Madrid 1969) 24–25.
Overall, one needs to remember that the forum was very large (153 m by 136 m on https://es.wikipedia.org/wiki/Foro_romano_de_Tarraco (last accessed Sept 2017)), and that very little of it has been excavated. Thus, some of these developments could have coexisted with a continued occupation of the plaza, in a classical style, confined to part of the great space. This can be supported by a dedication to Leo and Anthemius, found in the area of this square (RIT 100 = CIL 2.4109 = CIL 2–14–02, 947 = ILS 815 (AD 468–72)). However, it is thought not to be a dedication for a statue base, but rather for a fountain. The interpretation has been suggested based on a large hole cut into this block, around which the letters are set: LSA 1986. Nevertheless, the decay of a monumental stair tower, in the final quarter of the 5th c., is significant, as is the waste dump of ca. 440–50. The ceramic evidence can be used to indicate the start of decay, but decay and traditional occupation might have existed side-by-side in different parts of the plaza during the 5th c. As the ceramic dump on the forum is the earliest piece of decay evidence, I will take this as providing my date for the start of decay, in 440–50.
Dating summary (start of degradation, with spoliation of stair tower): range 450–500, midpoint 475, class Cs3 (associative finds?), publication 1/3.
Dating summary (start of decay, by which time spoliation of paving was well advanced in some areas): range 440–50, midpoint 445, class Cs9 (contextual ceramics), publication 3/3).
02HIS Complutum (Alcalá de Henares): There is no archaeological evidence available for the abandonment of the forum surface. However, the roofs of surrounding buildings collapsed sometime in the 5th c., as no finds have been produced from after this time. Firstly, a deposit under the basilica roof produced unspecified 4th–5th c. pottery and a coin of Theodosius II (AD 408–50), lying directly on the basilica floor: S. Rascón Marqués, La cuidad hispanorromana de Complutum (Cuadernos del Juncal 2) (Alcalà de Henares 1995) 99. Secondly, another building just off the forum, in a smaller plaza on the other side of the basilica, which served as a temple in its final 4th c. phase, has comparable evidence. A deposit under the roof (UE 119) with a great dish in ARS Hayes 59B [given as AD 320–420 in LRP/SLRP] and a bowl 37t and several [other] fragments of late TSHispánica (TSHT), given as of the middle of the 5th c., which could be in use up to approximately AD 500 [bowl TSHT ‘4.37tardia’ is given (oddly) as 450–ca. 450 in Paz, so agreeing with this]: S. Rascón Marqués, pers. comm.
Overall, the best basis for the abandonment of the forum is the dating of the ruin of the basilica (as this is part of the forum, whereas the temple faced onto a separate plaza that was much smaller). The collapse of the basilica roof can currently be given a rough contextual date, by combining the unspecified 4th–5th c. pottery with the Theodosian coin, which is the latest piece of published dating evidence. The coin provides a TPQ for the destruction, but we must extend the contextual date for the destruction to the whole of the 5th c., as this is the range given for the ceramics, so generating a TAQ of 500. Unfortunately, we do not yet have large areas of public buildings excavated within the city, against which to consider the development of the forum as typical or anomalous.
Dating summary: range 408–500, midpoint 454, class Cs9 (contextual ceramics), Cs8 (contextual coin), Cs7 (TPQ coin), publication 2/3.
02HIS Emerita (Mérida): The Colonial and Provincial Forum was abandoned in the 5th c.: P. Mateos, “La cristianización de la Lusitania (ss. IV–VII: Extremadura en época visigoda)”, Extremadura Arqueológica 4 (1995) (239–64) 245 (not seen).
Dating summary: Pending.
04VIE Rodez (Segodunum): This forum shows a complex history of refurbishment, followed by abandonment and spoliation. In the very last years of the 3rd c., the cellular rooms / shops on the south-east side of the forum were refurbished, after the area had been burned by a fire. There was then a domestic occupation of the ‘shops’, in two phases, before the site was partially spoliated: J. Catalo et al., “Le forum de Rodez: premiers résultats”, Mémoirs de la Societé Archéologique du Midi de la France 55 (1994) 11–58. These are also presented, with a focus on Late Antiquity, in J. Catalo et al., “Le Forum de Rodez (Aveyron) du IVe au VIIe siècle”, in L’Auvergne de Sidoine Apollinaire à Grégoire de Tours. Histoire et archéologie. Actes des XIIIes journées internationales d’archéologie mérovingienne (Clermont-Ferrand, 3–6 octobre 1991), ed. B. Fizellier-Sauget (Clermont-Ferrand 1999) 115–30. The phasing can be described as follows, on which I have relied mostly, except where the report of 1994 has extra details.
Fire: A fire damaged the south-east side of the Forum sometime in the late 3rd c.: Catalo et al. (1994) 32. The excavators believe that the totality of the finds reveal that the destruction dates from the very end of the 3rd c., although without specifying particular dating evidence from the destruction layers, which might have helped them reach this judgment (p. 39). The ceramic report is presented on pp. 32–38. Unfortunately, it is not organised in terms of individual contexts, but rather of wares across the whole phase, making it difficult to suggest a date for the fire. However, the report tells us that ceramics from the relaying of an adjacent street, before the fire, extend from the mid-2nd to the 3rd quarter of the 3rd c. (p. 35). Thus, despite the evidence not being fully published, I am inclined to defer to the excavators and accept a probable contextual date, of the late 3rd c., which is based on phase of development, in relation to the fire and subsequent occupation levels. This means the years 287.5–300 under the rules of this study.
Repairs: After the fire the east portico of the forum was repaired: Catalo et al. (1994) 32 describes the fire damage. A wall angle relating to the back wall of the forum portico was reinforced with reused material, whilst other walls were also repaired in this area [the south-east corner of the forum]. For simplicity’s sake, I will place this phase in the same date range as the fire, but before occupation phase 1.
Sewer interventions: After the late 3rd c. fire drains were cut into the sewer, which continued to function. The street also continued to function until the 5th c.: Catalo (1999) 116. I do not know the dating basis of this, although I suspect the stratigraphic relationships are derived from the lines of the drains as cut into the adjacent south-east forum rooms, whilst the terminal date is likely based on the end of finds from the street. I give the date for these works as 287.5–425, based on their position in relation to the fire and the end of the second occupation phase, which was followed by abandonment (see below), although soon after the fire is likely. We do not know the relationship of the works to occupation phase 1.
Occupation phase 1: There is evidence of what looks like domestic occupation in the row of three cellular rooms / tabernae on the south-east side of the forum. This occupation dates from the first half of the 4th c., according to the report. These changes did not affect the main area of the forum (portico, paving, exedras): Catalo (1999) 116. During this phase, a doorway leading into the southern shop was walled and a hearth was established here. Another hearth was installed in the large room to the south. These floors are posterior to the destruction fill of the end of the 3rd c., but contain homogeneous ceramics of the same period. These ceramics include: hemispherical forms “à bord rentrant et bandeau médian” (derived from Vernhet 1 and 2]); a mortar derived from Drag. 43 [of which Drag. 43 is given as AD 170–230 on potsherd.net]; imitations of ARS B (accounting for 45% of the ceramics); late sigillata from Lezoux (Drag. 37 [AD 70–230 on potsherd.net]; Ritt. 8 [Ritterling 8 given as AD 40–70 on http://potsherd.net/atlas/types/sigillata/form/RT8]; Drag. 35 [AD 70–230 on potsherd.net]; and a lamp à canal (firmalampen), given as of the end of the 3rd c., alongside various local and indistinct wares: pp. 116–118. However, the coarsewares represent a change to the 3rd c. deposits from the site, showing more grey fabrics of very good quality, with a glazed (lustré) surface (30%), in contrast to light coloured coarsewares known earlier (p. 118). The coins include (i) (from the dwellings) an imitation of Claudius II (given as of 270–90), and issues of Maximian (of 294); (ii) (in a cutting (tranchée) in the street) an imitation of Tetricius (given as of 270–90), Aurelian (of 270–75) and Dalmatius (of 335–37): p. 118–20. These finds are believed by the report authors to reveal an occupation sequence of the first half of the 4th c. In the absence of contextual publication we must defer to the excavators and accept a weak associative date of 300–50 for the occupation, based on the finds from these layers and a TPQ that gives time for the east portico repairs.
Occupation phase 2: In a later phase, another hearth was established beyond the wall that closes the paved plaza, in an area the report authors believe was public, not private space. The cellular rooms / tabernae on the south-east side of the plaza were also replanned (with a division wall being moved) and a new hearth was established. In deposits associated with this phase imitations of ARS B represent only a quarter of the sample. Furthermore, the grey fabric pottery described above now makes up the majority in the coarsewares, suggesting a chronological difference to the previous phase. There is also a small percentage of luisante pottery, with little coupes à bord rentrant close to the form of Lamboglia 3/8 of the end of the 3rd to 4th c. [150–250 in Atlante] This phase is dated to the second half of the 4th c. by relative chronology and by coins found in the floor of the (?rearranged) shop (p. 120). These coins are imitations of Tetricius (given as of 272–90), Gallienus (ca. 268), Constantius II (350–54) and Constantine II (337–40). The last deposits relating to occupation and the first abandonment layers have produced coins of the end of the 3rd and the beginning of the 4th c., of which the latest seems to be an imitation of Constantius II, minted in 337–61 (p. 122). The coarsewares reflect the groups previously encountered, except that pottery with ‘glazed’ edges disappears to the advantage of closed forms with a sandy fabric. There is also some stamped pottery from Languedoc, generally dated between 350 and the end of the 5th c. The only form clearly identifiable is a sherd derived from Rigoir 26 [Rigoir 26a / DSP 26a is 400–600 and 26b is 500/650 according to DCAMNO]. The report authors point out that this ensemble of forms is characteristic of the end of the 4th to the beginning of the 5th c. on the site (p. 122). Overall, unlike in other phases on the site, we have quite clear contextual information, as a group of coins is mentioned that comes from a floor of the tabernae. The floors are slightly earlier in date than the ceramics, suggesting that they are finds incorporated in construction rather than trampled in as part of use, thus providing a contextual date for the shop floors of 350–75. The latest coin (beginning in 350) also provides a TPQ for occupation phase 2. Ceramic finds from occupation phase 2 provide a weak contextual TAQ (as use finds) of the 25 year period after 400, based on their last finds being Rigoir 26.
Of the functional character of occupation phase 2, there is no indication of anything but domestic occupation in the first late phase of shops, whereas the second phase (with replanning) has numerous finds associated with bone-working (pins, needles and tokens) (pp. 122–24).
Of abandonment, the pottery from the last layers of occupation, “destruction” [of which ‘spoliation’ seems to be the correct meaning] and abandonment is reported as being very homogeneous. Aside from poorly executed grey wares, belonging to late ceramics from Provence, there are also sherds inspired by Rigoir 6 and 16. The assemblage corresponds to forms widely-known for the 6th c. in the Massif-Central / Languedoc and Provence region. There are also a few sherds of grey or black vases biconques typical of the north of Gaul in the 6th c., which are related to Frankish incomers. Metalwork of 7th c. date is also present, but I will not go into the details of this. The excavators suggest that we have here an abandonment of the forum, beginning in the 5th c., but that the spoliation of the structure dates from the 6th to 7th c.: Catalo (1999) 125–26 (see below). Unfortunately, they do not explain what physical features might indicate abandonment, such as a silting of the forum surface. However, they mention on p. 124 that traces of occupation practically come to an end in the 5th c. in the forum. We can at least note that the occupation period 2 (in the east portico, above) seems to extend to 400–25 based on the ceramics found. This same ceramic evidence can give a TPQ for the beginning of abandonment of AD 400 for the forum.
Occupation phase 3: A final phase of occupation is indicated by a dwelling established at the centre of the east portico, in a different excavation north of that described thus far, reported in Catalo (1994) and (1999). This dwelling was made up of a mortared wall of reused material and a sequence of three mortar floors, before the pier against which it was established was robbed of its stones [the word épierrement is used], like other walls on the site. The occupation levels [presumably of the dwelling in the portico] produced fragments of unspecified glass and unspecified ceramics, dating from the ‘same century’ [here the 6th c.], suggesting a rapid succession of occupation and destruction: J. Catalo, “Rodez: du forum antique au couvent des Jacobins”, Aquitania 8 (1991) (161–86) 175–77. This dwelling indicates that the portico had been at least partially privatised, probably with a good number of its portico intercolumniations being blocked. I accept the full range of the ceramics given (as no wares are specified) as providing an associative date for the occupation (and construction of the dwelling) of the whole of the 6th c. The TAQ of 600 derived from the subsequent phase does not change this.
Of spoliation, the excavators [still Catalo (1991) 175–77] wish to associate the robbing of the dwelling with a ditch that runs across the north side of the site, from east to west, into the adjacent sewer, and may be a drainage feature. Another ditch of the same nature is present on the far side of the sewer [so to the east]. The excavators further note that the spoliation is more severe to the south than to the north of the east-west ditch, as if the cutting of the ditch had led to the spoliation of walls discovered. This theory is suggested by the fill of the ditch, which is identical to that backfilling the robbing trenches of the adjacent walls. The fill from the two drainage ditches is very homogeneous and has produced DSP tardive [DSP range in general is 370–650 in DCAMNO] and rims of very distinctive grey ceramics (found across the Massif Central and comparable to céramique paléochrétienne grise et orangée from Languedoc, which looks very similar to DSP, of which the established alternative name is céramique estampée grise et orangée), alongside glass of a ‘Germanic’ type [base of cup carchesium; rim of small cup glassschalen of level Böhner III 525–600 for the north of France, citing K. Böhner, Die fränkischen Altertümer des Trierer Landes (Berlin 1958) (not seen)]. This suggested to the excavators a date for the ‘drainage’ ditches and spoliation of the second half of the 6th c. AD, although it would seem that the glass is the last dated find specified. Until I have further publication evidence, I accept the full range of 550–600, given by the report, for this earliest recorded incident of spoliation, as ceramics are clearly involved. They are not able to name these wares, reflecting in part a lack of development of studies of local ceramics in their region, but also a lack of precision about pottery, which the later report of Catalo (1999) does not show. Elsewhere, spoliation was very varied in its character. The forum paving was not stripped, and some walls were left untouched, whereas in other places the robbing was intense: Catalo (1999) 125–26. These remarks seem to concern the south-east area reported in this article but which might possibly include a re-evaluation of the area to the north reported in Catalo (1991), although this is not made clear.
Dating summary (fire): range 287.5–300, midpoint 293.75, class Cs3 (associative, phase of development), Cs3 (associative, finds), publication 2/3.
Dating summary (repairs to portico wall and adjacent rooms): range 287.5–300, midpoint 293.75, class Cs3 (associative, phase of development), Cs3 (associative, finds), publication 2/3.
Dating summary (drains to sewer) (2): 287.5–425, midpoint 356.25, class 0, publication 1/3.
Dating summary (occupation phase 1, not a building phase): range 300–350, midpoint 325, class Cs3 (associative, finds), publication 2/3.
Dating summary (shops replanned): range 350–75, midpoint 362.5, class Cs8 (contextual coins), publication 2/3.
Dating summary (occupation phase 2): range 350–425, midpoint 387.5, class Cs7 (TPQ, coins), Cs7 (TPQ ceramics), Cs3 (associative, finds), publication 2/3.
Dating summary (beginning of abandonment and spoliation): range 400–425, midpoint 412.5, class Cs7 (TPQ, ceramic), Cs3 (associative, phase of development), publication 1/3.
Dating summary (occupation phase 3, construction of dwelling in portico): range 500–600, midpoint 550, class Cs3 (associative, finds), publication 2/3.
Dating summary (‘drainage ditches’ and related spoliation): range 550–600, midpoint 575, class Cs9 (contextual ceramics), Cs8x (contextual, other), publication 2/3.
04VIE Arles (Arelate): The late evolution of the Forum of Arles is a controversial topic. Archaeological evidence appears to contradict the account of Sidonius, Ep. 1.11.7 (AD 461) of an architecturally intact forum for the city, with surviving columns, which was still a setting for important social encounters. The account of Sidonius (transl. of Dalton 1915), describes his visit to the forum at a time when he was accused of writing a satire against the Emperor Majorian, who was residing at Arles:
“The next day I paid my duty to the emperor, and went down to the forum, as I always do. As soon as I appeared, the conspiracy was at once confounded, being of the sort which, as Lucan says, dares put nothing to the touch. Some fell cringing at my knees, abasing themselves beyond propriety; others hid behind statues or columns to avoid the necessity of salutation; others, again, with looks of affected sorrow, walked closely at my sides. I was wondering all the time what might be the meaning of this excess, first in insolence and now in abasement, but was determined not to ask, when one of the gang, put up, no doubt, to play the part, came forward to exchange greeting. We talked, and incidentally he remarked: ‘You see these people?’ ‘I do indeed,’ I answered, ‘and I may say that their proceedings astonish me as much as they impress me little.’ To which my kind interpreter rejoined: ‘It is in your quality of satirist that they show this fear or detestation of you.’ ‘How so,’ I cried, ‘on what grounds? when did I give them the excuse? who detected the offence? who brought the charge and who the proof?’ Then, with a smile, I continued thus: ‘My dear sir, if you don’t mind, oblige me by asking these excited persons from me, whether it was a professed informer or spy who got up this imaginative story about my writing a satire. If they have to make the inevitable apology later, it will be better for them to give up this outrageous behaviour at once.’ No sooner had he conveyed the message, than they all came to offer their hands and salutations, not man by man, and with decorum, but the whole herd with a rush. Our Curio was left all alone to breathe imprecations on the base deserters, until at fall of evening he was hurried off home on the shoulders of bearers gloomier than mutes.”
The text is interesting as it describes not only columns and statues, but also traditional social interactions, with groups of clients keen to advertise their allegiances on the plaza, alongside the greeting of patrons and the presence of a litter carried by bearers. These are all features of public life attested in 4th c. Antioch / Constantinople by Chrysostom and Libanius (see main chapter). It is worth remembering that only a tiny area of the forum of Arles has been excavated, before we turn to the archaeology.
Firstly, in the gallery / cryptoporticus of the northern portico of the forum (eastern end) was found a dump of marbles including a shield and head of Augustus, cornices and other fragments of architectural decoration. These marble pieces were mixed with pottery. The ceramics consist of two pieces: firstly, a sherd of Luisante pottery, that Heijmans identifies as being 5th c. in date (produced around Monmoiron in the Vaucluse) and, secondly, an amphora which he identifies as a spatheion of 5th c. date: M. Heijmans, “Nouvelles recherches sur les cryptoportiques d’Arles et la topographie du centre de la colonie”, RANarb, 24 (1991) (161–200) 196–97, esp. 196 n. 109. Heijmans refers here to F. Benoit, “Le sanctuaire d’Auguste et les cryptoportiques d’Arles”, RA 39 (1952) 31–67, who notes on p. 64 that the above pottery comes from the lowest level of the dump, alongside the marbles. See also J. Latour, “Le sanctuaire d’Auguste et les cryptoportiques d’Arles”, RA 42 (1953) 42–51, with ceramics mentioned on p. 51. A different layer of ceramics, found above the marbles, described on p. 44 with p. 48 fig. 6 and p. 50 fig. 8 (section drawing), does not seem to be pertinent. Heijmans interprets the lower dump as reflecting the decay of the forum area, as the marble pieces perhaps indicate the presence of a lime kiln. It is also possible that it represents a marble recycling store, of pieces ready to be reused or cut into veneers. The possibility that the galleries of the portico were being used for storage is evoked by Heijmans. He notes the undated appearance of mortises in the pillars of the galleries, which might relate to such an occupation. Heijmans also mentions a floor in the south portico, which he interprets as a dump, without giving further details: Heijmans (1991) 196.
Secondly, more significantly, there is the appearance of ‘parasitic’ buildings, over the surface of the forum itself. This evidence was discovered in a small trench at the Town Hall, which revealed 30 m square, about 1% of the surface of the forum, at its south edge. These ‘parasitic’ buildings are dated from the end of the 4th or the beginning of the 5th c.: J.-M. Rouquette and Cl. Sintès, “Arles: D’hôtel de ville”, Gallia Informations (1987–88.2) 233–34; J.-M. Rouquette and Cl. Sintès, “Arles, hôtel de ville”, Notes de la Direction PACA 4 (1987) 77–78 (not seen). The report given in Rouquette and Sintès (1987–88.2) 233–34 with fig. 47 (plan) and fig. 48 (photo) describes a wall composed of small blocks and large pieces of reused material, which was built at a 90 degree angle across the dismantled stylobate. The new building had a beaten earth floor, under which was found a number of architectural fragments serving as a make-up layer. In this make-up layer, and just above it, were found ceramics indicating that the ‘parasitic’ building was established in the first half of the 5th c. The following ceramics were described: ARS D of Hayes 61A [ca. 325–380 in LRP with SLRP], 61B [ca. 380–450 in LRP with SLRP, although Bon has all variants related to 61B as 400–500], 62 [ca. 350–425 in LRP], 67 [ca. 360–470 in LRP] and 91A [400–450 Bon], with a high proportion of ‘ceramique luisante’ from the workshops of Portout and Conjux. Although I cannot identify the form of the last ware, it seems likely that ceramics relating to the second period of production of Portout in the 5th c. is implied: C. Raynaud, “Céramique luisante”, Lattara 6 (1993) (504–10) 504. As the ‘first half of the 5th c.’ date is offered by a competent local study of pottery, I am prepared to defer to it, with regret that the latest ceramics are not made clear. I take this evidence as providing a more accurate impression of the start of degradation on the forum than the deposit from the cryptoporticus. The level of deference is not very high, as Hayes 91A suggests that the ‘parasitic’ building dates no earlier than 400.
Overall, the presence of the ‘parasitic’ building on the paving is evocative of degradation, especially because it implies the demolition of the portico. However, the size of the trench and its position against the edge of a very large plaza (of just over 84 m by 23 m, hand-measuring off the map of Heijmans (1991) 162 fig. 1) should not be taken as excluding the continued existence of a monumental forum at Arles. In other cities encroachment and monumental improvement can be seen coexisting, notably at Cyrene, where different parts of the agora had very different histories. Furthermore, cities often had more than one forum, so we cannot be sure that the same plaza is referred to by Sidonius in his letter. Nonetheless, I am happy to date degradation on the forum excavated here to the first half of the 5th c. based on contextual ceramics.
Dating summary (of start of degradation, of at least part of the forum): range 400–50, midpoint 425, class Cs9 (contextual ceramics), publication 3/3.
04VIE Vienne: The forum became home to deer by 474, as part of the general decay of city: Sid. Apoll. Epist. 7.1.3 (AD 474). There is no TPQ possible for this process. However, it would make most sense, as a personal observation of something new, if the process of decay began within the adult lifetime of Sidonius (born AD 430). Thus, I would suggest 444–74 as a reasonable range.
Dating summary: range 444–74, midpoint 459, class x (historical text), publication 3/3.
05ITA Verona: Degradation can be seen as beginning in the period 383–408. This degradation was only partial and does not amount to abandonment. Degradation can be seen in two areas (i) and (ii) from archaeology, which must be set against another area (iii) that shows less transformation, plus more positive testimony derived from epigraphic (iv) and literary texts (v).
(i) The capitolium excavations have revealed a destruction layer, including much masonry and large tile fragments, which was found covering the paving of the pedestrian walkway around the west side of the temple enclosure. This related to the destruction of the triporticus wall surrounding the temple. The layer included mostly 3rd to 4th c. material. However, the latest material consisted of a coin (AE4) of 383–403, ARS Hayes 67 [ca. 360–470 in LRP], 50B [ca. 350–400+ LRP], 61B [ca. 380–450 in LRP with SLRP, but Bon 400–500, which I now follow], and amphorae LRA4 [300–700 on RADR] and Keay 52 [350–700 RADR], alongside glazed pottery and soapstone wares. Finds of 5th c. date were also noted in the surface, but not specified in the report: G. Cavalieri Manasse, “Gli scavi del complesso capitolino”, in L’area del Capitolium di Verona: ricerche storiche e archeologiche, ed. G. Cavalieri Manasse (Verona 2008) (73–152) 112–14. I take this deposit as providing a contextual date for the destruction of the pedestrian walkway / triporticus wall of the temple in the 25 years after the start of its last dating indicator, giving a range of 400–25.
(ii) The west portico of the forum was invaded by ‘hovels’ sometime in the 5th c., after a phase of spoliation: Cavalieri Manasse (2008) 115, drawing on G. Pietra, Dalla città romana alla città altomedievale: l’analisi della ceramica di Verona—Piazza Erbe 27 (Tesi di specializzazione, University of Padua 1999–2000) (not seen). Without sight of the report, I was not able to establish the dating basis for this. However, the absence of architectural fragments in the robbing layers of the temple [i.e. the capitolium], in contrast to those in later layers, suggests that the architectural structures above these ‘hovels’ were intact.
(iii) Excavations elsewhere in the plaza (a report cited as ASAV, Verona, Piazza erbe 21, 1984, not seen) revealed less evidence of transformation of use: Cavalieri Manasse (2008) 115. The whole area was at some point affected by fire destruction, with resulting calcinations, which has been suggested to relate to the fire recorded by Paul. HL 3.23, in the 580s, which has possibly been identified archaeologically elsewhere in Verona (at via Dante): see appendix A8b.
(iv) In contrast, an inscription records the erection of a statue brought from the capitolium into the forum, at the command of PLRE 1.662 Val. Palladius 19, consularis, under the Augusti Gratian, Valentinian and Theodosius (so 379–83): CIL 5.3332 = ILS 5363. This inscription suggests that at this time the plaza was still a representational area.
(v) There is also a literary description of the forum of Verona provided by Versus de Verona 4 (MGH Poetarum Latinorum Medii Aevi vol. 1) (Berlin 1881) 119–22). This text (of ca. 800) describes ‘a wide and spacious forum paved with stones, at each of its 4 angles stands a great arch’. This would suggest that Late Antiquity had been a period of architectural survival as much as of change. It is possible that the Versus refers to a second forum, but given the small areas excavated on the main forum it is not possible to rule out the accuracy of the Versus, which is a clear description of a forum that had lost its redundant political and pagan religious buildings, but retained classical monumental elements, still popular in Late Antiquity.
Overall, although the capitolium has the best evidence, we have to consider that this temple might have been removed from the plaza as part of an act of selective editing: its history cannot be taken as representative of the plaza as a whole. However, we can at least derive an associative TPQ for the plaza from the destruction layer (of 400–25) of the capitolium, because it is unlikely that the main plaza decayed before the temple. In terms of TAQ for the degradation of the forum, the west portico seems to provide some basis; it had already been spoliated prior to its invasion by hovels in the 5th c., meaning that the degradation of the plaza must have occurred prior to 500. Unfortunately, I was not able to establish the basis of the dating for the west portico, meaning that the forum date must be partly classed as ‘pending’.
Dating summary (start of degradation of forum): range 400–500, midpoint 450, class Cs9 (contextual ceramics), publication 2/3. Pending.
Dating summary (temple destruction): range 400–25, midpoint 412.5, class Cs9 (contextual ceramics), publication 3/3.
05ITA Aquileia: The forum, which had seen repairs in the 4th c., was excluded from the city by the last major wall circuit in the city, which has an unusual saw-tooth profile. Evidence for the abandonment of the forum comes from 4 sources: (i) The fire destruction of the forum basilica produced burnt coins, including 4 of Valentinian II [375–92], one of Gratian [367–383], and one of ‘Theodosius’ [379–95 if Theodosius I, which is likely as Theodosius II was an eastern Augustus]: I. Giacca, “Monete rinvenute nello scavo degli ambienti tardo-romani e della basilica forense a sud-ovest del foro romano di Aquileia”, AquilNost 51 (1980) 164 and P. Lopreato, “Aquileia: lo scavo a S-O del foro romano. Gli ambienti tardo antichi e la basilica forense. Relazione delle campagne di scavo 1977–1979”, AquilNost 51 (1980) (21–54) 50 n. 14 [N.B. the only later coin is of Valentinian III, but is from a post-destruction wall]. See appendix X1a; (ii) Certain memorable civic inscriptions were placed face down in the plaza, rather than being reused, suggesting an organised and respectful abandonment of the forum: C. Zaccaria, “Permanenza dell’ideale civico romano in epoca tardoantica: nuove evidenze da Aquileia”, Antichità Altoadriatiche 47 (2000) 91–113. For commentary, see: C. Sotinel, Identité civique et christianisme. Aquilée du IIe au VI siècle (BÉFAR 324) (Rome 2005) 251–57; (iii) A spoliated inscription was recovered from the saw-tooth city wall, relating to the three emperors Valentinian, Theodosius I and his son Arcadius (so dating from 382–91): G. Brusin, Gli scavi di Aquileia (Udine 1934) 84–85, no. 8; (iv) A tower attached onto an earlier fortification includes a significant amount of spolia from the forum complex. The tower has been argued to be similar in form to those on the saw-tooth wall, which excludes the forum, and so could be of the same date, but this is conjecture.
Overall, we can perhaps reconcile these different threads of evidence, if we envisage a fire destruction of some buildings, followed by organised abandonment, then the building of a new city wall. This is the only order of events that seems to make sense: the civil basilica would not have been left standing beyond the new city walls after they were constructed, as it lay north of the line of the wall and would have obstructed artillery fire. It is tempting to place the fire in relation with the well-documented sack of the city by Attila in 452 or as a result of an attack by Alaric in 410. Although such correlations between written and archaeological sources are dangerous, it does seem to be plausible to link this fire to enemy action in the case of the Huns. The destructive effect of a Hunnic sack seems well confirmed by archaeology at Nicopolis ad Istrum, for which see appendix C10a. However, at present it is best to allow a contextual coin date from the basilica to set the chronology of abandonment. Dated comparanda for the saw-tooth wall suggest that it belongs to a later period, perhaps that of the Justinianic reconquest from 535 onwards, which is the date preferred by a recent comparative study: S. Groh, “Forschungen zur Urbanistik und spätantik-byzantinischen fortifikation von Aquileia (Italien): Bericht über die geophysikalischen prospektionen 2011”, ÖJh 81 (2012) (67–96) 91–93. Parallels suggested for the saw-tooth wall include: Thessalonica, dated by brick stamps to the later 3rd or the 5th c., before 491; Amorium, dated to sometime after 487, based on dendrochronology, references I do not want to go into here. In the presence of this new dating for the fortification, it is best to stick with the fire destruction evidence alone in dating the abandonment of the forum.
Dating summary (abandonment of forum, after fire): range 379–404, midpoint 391.5, class Cs3 (associative, phase of development), Cs8 (contextual coins), publication 3/3.
06ITS Nora: The decay of the forum of Nora can be seen in the number of different elements, which can be taken as representing a phase of decay, then encroachment / reoccupation: A.R. Ghiotto, “Il complesso monumentale del Foro”, in Nora. Il foro romano. Storia di un’area urbana dall’età fenicia alla tarda antichità (1997–2006). Vol. 1. Lo scavo, ed. J. Bonetto (Padua 2009) (245–373) 361–66.
Of phasing: (i) In the west portico (restored in AD 394–419, see appendix K4a), the paving of the northern section was stripped; three hearths were set against the back wall and in the north-west corner of the portico; the wall-plaster was partially removed (prior to being burned by the hearths); (ii) In the east portico the paving was stripped; (iii) In the west portico an encroaching room was established within the northern section, by closing the colonnade and building a wall including a doorway across the width of the portico, all using reused materials; the structures were established after a layer of light brown earth (US 11531) was laid down to form a new level, containing many tiles and much mortar; this level was cut by rubbish pits; (iv) In the east portico an encroaching room was established, after the stripping of its paving; this encroachment only partially survives, as one section of intercolumniation blocking wall (in reused material) and a lateral wall, found adjacent to it, which was built across the portico; (v) The forum paving was eroded [but no spoliation is proven] and was covered by abandonment layers, whilst a forum cistern also filled up; (vi) Structures were built on the forum paving, which have only been detected in the form of mortar traces on the paving in the north-east corner of the plaza.
Of dating, there is only evidence from the western portico (northern section). Here the light brown earth layer forming the new level for the encroaching room (US 11531) produced 11 sherds of 2 forms, believed by the excavators to relate to the last quarter of the 4th c.; these were Hayes 62B n. 15 [62B is ca. 350–425 in LRP] and Lamboglia 9A [confusingly 187.5–412.5 in Atlante, but given as equivalent to Hayes 27 in LRP, which would make it 200–300 in Bon], and 2 sherds of comb-decorated ceramics, which see their widest production between the 4th and the 8th c. (see C. Nieddu, “La ceramica medievale”, in Nora. Il foro romano. Storia di un’area urbana dall’età fenicia alla tarda antichità (1997–2006) Vol. 2.2—I materiali romani e gli altri reperti (Padua 2009), edd. J. Bonetto, G. Falezza and A.R. Ghiotto vol. 2 (757–61) nn. 1–2, who does not use it to contradict the wider dating of the context. In the same context 11531 and in 11539 and 11552 (related layers of dark brown earth levelling up an uneven part of 11531) were found 11 coins, of which the most numerous and recent date to the 2nd quarter of the 4th c. (M.G. Pavoni, “Le monete”, in Nora Vol. 2.2 (as above) (871–80) nn. 14–15 (of 336–37 of Constantine; 330–33 of Constantine for Constantius II) and 17–22 (347–48 of Constans; 335–37 and 346–47 and 347–48 of Constantius II and 335–41 twice of Constantine I and his family)): p. 368. Thus, we have the levelling layers posterior to the robbing of the west portico, from which the last finds have a start date of ca. 350, under LRP’s dating of Hayes 62B.
However, these levelling layers are already posterior to the rebuilding of the west portico, placed in the period 394–419 based on contextual coins (the last being an issue of Arcadius, of AD 394–95) (see appendix K4a), meaning that the finds of the levelling layers have little value. The report writer would like to place both the robbing and encroachment of the portico after 450, as he has attributed the previous phase to 400–50 [rather than 394–419, which is what I use]. Whilst this is plausible, it is not possible under the rules of this study. I can accept that decay began (with stripping of veneers etc.) at least 25 years after the rebuilding of the portico, based on allowing reasonable time for this phase of development, so after 419. But the TAQ is less certain. The finds from the light and dark brown levelling layers were clearly somewhat residual and do not give us a clear date. The absence of any finds from after 394, from the west portico area, makes one think that decay began no later than 25 years after the end date for this last phase of renewal, so 444. Then encroachment can be placed after this in a further 25 year period. This is a poor date, given the crude phase of development dating I have adopted, but is reasonable for an excavation that has not produced any finds with a start date of after 394.
Overall, in the absence of other evidence, we can tentatively associate the other elements of decay recorded around the plaza with the sequence recorded in the west portico. Finally, we might put the reoccupation of the plaza (with encroaching buildings) into an arbitrary phase of a further 25 years length after this. I am here indulging in arbitrary phasing, therefore it makes for a poor date.
Dating summary (spoliation, beginning of decay): range 419–44, midpoint 431.5, Cs3 (associative, phase of development), Cs8 (contextual coins), publication 3/3. Poor.
Dating summary (encroachment of west and east porticoes) (2): 444–69, midpoint 456.5, Cs3 (associative, phase of development), Cs8 (contextual coins), publication 3/3. Poor.
06ITS Egnatia: A roughly square room was built over the south-west portico and part of the forum paving, whilst a series of new rooms was built over the ruins of an adjacent north-west portico: R. Cassano, “La vicenda urbana di Egnazia ridisegnata dalle recenti indagini”, Fasti Online (2009) 1–20 (p. 5): http://www.fastionline.org/docs/FOLDER-it-2009-161.pdf (which provides details of ceramics found stored in artisanal workshops inside these rooms on p. 6). This development could be considered an example of unaesthetic degradation of a plaza that continued to exist as an open space.
Of phasing, the north-west portico was ruined, with new rooms built in and over it. Fragments of the portico, especially columns and architraves, served as building material for these rooms, implying that it was now ruined. The plan on p. 4 fig. 5 also indicates that the south-west portico was ruined by the time another room was built over it. Furthermore, the thresholds of the buildings built in and over the north-west portico include one that was set ca. 30 cm above the paving level. This detail suggests that the forum plaza at that time was covered in silt or some other material: L. Lavan site observations, April 2016. Thus, we have a plaza in decay, but no abandonment. The new units had floors of beaten earth.
Of dating, the report tells us (pp. 5–6) that the rooms were established after the decay of this area, coinciding with the deposition of coins of Valentinian (I take this as Valentinian I and so 364–75) and his successors, which the excavator thinks indicates decay at the end of the 4th c., so implying that the coins must be of Gratian (375–83), Valentinian II (375–92), Theodosius I (379–95), and possibly Maximus (383–88) and Eugenius (392–94). We do not have many details on the context of the coins. Also, we cannot know from the report what is the start date of the latest definitely identified coin, so I take it as being 375, which is the earliest date when a successor to Valentinian I could have issued a coin. We cannot calculate a proper contextual date for this information, but the 25 years from 375 to 400 seems a reasonable period of deposition of this assemblage, given that the excavators do not indicate the coins extending into the early 5th c.
Of function, the rooms seem to have served a commercial purpose, as balances and many coins were found here. This is interesting, as it indicates that the plaza (which was not built over, apart for one room) continued to host trading. The finding of a great quantity of hooks and needles of varied types, related to the repair of nets, suggests that chandlery was practiced here. Storage vessels set into the ground suggest food storage, as described in printed boards seen in the site museum by L. Lavan on site in April 2016. Weights for nets were also visible, displayed alongside hooks and needles, in the site museum.
Dating summary (establishment of new rooms): range 375–400, midpoint 387.5, class Cs8 (contextual coins), publication 2/3.
07AFR Thamugadi (Timgad): Slots were cut into the forum paving, suggesting that a change of function occurred at a late period: S. Roskams, “The urban transition in the Maghreb” in Early Medieval Towns in the Western Mediterranean, ed. G.P. Brogiolo (Mantua 1996) (43–54) 46.
Dating summary: undated.
07AFR Carthage (Admiralty Island plaza / ‘Maritime Agora’): This round plaza was built ca. AD 200 to cover the island at the centre of the Circular Harbour, where the Punic ship sheds had once stood. The plaza measured ca. 92 m from portico back wall to back wall, or ca. 105 m across, from one external colonnade stylobate to another (hand-measured from various plans in Hurst (1979) below). Based on diameter of ca. 92 m the portico’s internal perimeter length, along the back wall, should be ca. 289 m. It is a major monumental complex, bordered by a circular portico, with external and internal colonnades, separated by an internal wall. It encloses a vast space, in the centre of which were two buildings (Temple 2 and building 3). The plaza had one major entrance, through a monumental arch, on the north side. This led off a causeway that connected the island to the mainland. The complex may be the platea maritima / ‘Maritime Agora’ of the city (see appendix S4), which saw two phases of encroachment: the first in the 3rd to 4th c., the second in the later 6th c., separated by two substantial phases of renewal. Unfortunately, the British excavations of this area have not yet been fully published, so I rely here on the interim report of H. Hurst, “Excavations at Carthage 1977–8. Fourth interim report”, AntJ 59 ((1979) 19–49. This provides a rough summary of the phases of the island. These phases are based on level theories (of common heights uniting different work), masonry types and some excavated finds, drawing also on the excavations of earlier scholars.
The first phase of encroachment involved the construction of fairly well-ordered, though unmonumental, secondary buildings and structures within the complex: Hurst (1979) 39. None of these developments appear to add to the aesthetics of the plaza, so should be considered to represent degradation. They can be described as follows: (i) Firstly, and most obvious of all, is an area on the north-east side of the plaza, towards the portico, with at least two successive periods of stone and earth foundations, for mudbrick structures (buildings 4 and 5), which “may all date to the fourth-early 5th centuries” [no reason given]. These were cut by pits containing (unspecified) late 5th c. pottery (p. 41). (ii) Secondly (back on p. 39), there is an area of rubble and earth (broken up mortar-rubble masonry with occasional fragments of painted plaster, roof tiles and mosaic) that may represent a robbed-out building associated with [unspecified] 5th c. pottery. Unfortunately, we cannot know whether this pottery relates to the occupation or robbing of the hypothetical structure. Both developments are shown on a map on p. 38 fig. 8. It is possible that they may be associated in time with a deposit of ostraca, found on the north side of the square, close to the entrance, detailing the work of Felix the ‘mensor olei fori Karthag[inis]’ of AD 373 (dated by consular dating), which appears to describe quantities, perhaps of oil, brought in and out of the harbour on different sailings. However, on present evidence, I do not think it is possible to make a chronological connection between the encroaching buildings and this evidence. Rather, we should just say that the encroaching buildings were secondary to the primary monumental design of the plaza ca. 200, and probably date to after 250, on account of the aesthetic disjuncture that they represent. They seem to have been demolished in the late 5th c., after a sequence of at least two sub-phases. The upper limit for dating the first phase of encroachment is provided by finds of the subsequent phase, a period of renewal, dating from the first quarter of the 6th c. (see below). Thus, a loose dating of 250–525 is necessary, using the next period as a TAQ, even if one might speculate that the encroaching buildings belong in the 5th c., being established after the end of the annona for Rome, at the time of the Vandal conquest of the city in AD 439.
The first phase of ‘renewal’ for the plaza (shown on the map on p. 40 fig. 9) comes in the mid-5th to early 6th c. It does not amount to very much: (i) Temple 2 was obliterated. A small part of the structure was retained to serve as a covered well, to which were associated building levels cut by pits containing unspecified late 5th to early 6th c. pottery. A surface metalling of gravel and potsherds, of mid-5th to early 6th c. date, lay above these (p. 41). (ii) Building 3 may also have been demolished, as rubble covering steps on the west side of the building contained pottery and lamps of late 5th c. date (first claimed as coins and lamps of late 4th c. date in H. Hurst, “Excavations at Carthage 1974, first interim report”, AntJ 55 (1975) (11–40) 27). (iii) In the north-east part of the plaza, above the fill of the pits cut into the demolished mudbrick buildings, graves were cut. On these burials, see A. Leone, “Le sepulture nello spazio urbano a Cartagine tra V e VII secolo d.C.”, AnTard 10 (2002) (233–48) 244; S. Stevens, “Sépultures tardives intra-muros à Carthage”, in L’Afrique du Nord antique et médiévale: monuments funéraires, institutions autochtones, ed. P. Trousset (Paris 1995) (207–17) 211. (iv) A massive dump of material was laid down on the north and south-east edges of the island, against the Roman quay wall, which might be the foundation for a (lost) late quay. The dump was uniform, indicating that it was laid down in a short period. The dump contained pottery and lamps of the first quarter of the 6th c. It has to be admitted that the metalling (with mid-5th to early 6th c. pottery) and the quay dump (with material of the first quarter of the 6th c.) are the only parts of this phase that represent positive acts of construction. However, the plaza was clearly opened up again. If we accept that the works described do belong together in a single phase, then we can give them, under the rules of this study, a date from (iv), as the latest deposit, of 500 to 525, being a 25 year period, which coincides with the full date range provided for these ceramics.
As a post-script to the first phase of ‘renewal’, it is worth noting that Hurst thinks, as a guess, that burials within the internal colonnade, once suggested as being 5th c., might actually belong in the 7th c., putting them into the second phase of encroachment: H.R. Hurst, Excavations at Carthage: The British Mission, Vol. II.1. The Circular Harbour, North Side: the Site and Finds other than Pottery (Oxford 1994) 114. The burials described above might also have been part of the second phase of encroachment. If so, the first phase of renewal would look more positive and designed to restore the public character of the plaza.
The second phase of ‘renewal’ of the plaza involved a much more comprehensive monumental rebuilding. This phase is seen by Hurst (1979) 41–3 as belonging to the mid-6th c. It was composed of 4 elements, which were closely connected in structural terms: (i) At the edge of the harbour, great quantities of earth were tipped, covering over previous pottery dumps. This earth formed a hard rubbly band around the island, indicating an extension of its surface. However, the interface with the sea itself (perhaps a late sea wall) does not survive. (ii) The causeway linking the island to the mainland seems to have been rebuilt, with a line of foundation material (of rubble and earth robbed from the previous structure), with a grey mortar covering over the ruins of the broken vaults. The causeway may have been narrower than the previous Roman construction, as the grey mortar extends over a width of ca. 8 m rather than the ca. 12 m, as it did for the Roman causeway. (iii) A monumental entrance was constructed, at the end of the causeway, where it reached the exterior of the island / round plaza. This stood in place of a Roman arch, which had existed there previously (the report tells us nothing about its demolition). Four massive rectangular foundations were built here of “mortared rubble construction with a high content of fine-grained micritic limestone, mixed with some reused stones of various types bonded in a charcoal-flecked grey mortar”; this masonry contrasted sharply to the Roman period mortared rubble structures, which had different sandstone types and whitish or yellowish mortars. (iv) A new portico was established: a circuit wall running round the island and a series of square bases on the inner side were built in a comparable limestone and grey-mortar masonry type (detected in 5 places on the plan). These seem to represent a portico to replace the earlier inner portico of the plaza, although the new portico back wall was built slightly forward from the line of its Roman predecessor, by perhaps ca. 0.5 m or so, judging from the plan on Hurst (1979) 49 fig. 9. Hurst provides no measurements for the features that he describes, but they can be estimated, by hand-measuring from the plan on Hurst (1979) 49 fig. 9. From this we can see that the portico measured ca. 3.5–4 m deep from back wall to plaza side of colonnade (represented by square bases, perhaps for piers as much as for columns; this is not clear from the text). Of dating, the 4 structures in this second period of ‘renewal’ can be tentatively dated: a cistern within the island was built in the same charcoal-flecked mortar, although with slightly different stone types. This cistern had mid-6th c. pottery in its construction trenches, which should indicate a date in the 25 years around 550 under the rules of this study, using the full range of the pottery date provided, as we do not have individual wares. This date, based on phase of development and catch-all masonry theories, is not contradicted by other features with this mortar style, in the wider sequence of the plaza.
Of the late monumental entrance, the 4 piers represent something of a problem. They are not entirely aligned with each other, suggesting perhaps a tetrastylon or tetrakionion rather than a tetrapylon, although this might by the result of earth movements, given the water in the ground here and their proximity to the edge of the harbour. Considering the irregularity of other new structures in the plaza, one might also suggest that they were just not built in a very regular manner.
Of dimensions, we can say that the monument is inscribed within an almost rectangular quadrilateral. This measured approximately 7.8 m on both long sides (facing the land or facing the plaza), but 5.9 m on the east side and 5.2 m on the west side. The piers each measure individually ca. 1.75 m east-west by 1.35 m north-south, although the south-west pier is not complete and is thus (perhaps erroneously) marked on the plan as measuring ca. 0.18 m less in a north-south direction (I will disregard it in favour of regular measurements suggested by the other three piers). The east-west separation between the piers is ca. 4.3 m in both cases, whilst the north-south separation (not important, as traffic did not go through here) was 3.2 m on the east side and 2.5 m on the west. The piers are arranged so that the south-east example is conjoining the corner of the portico back wall. However, the other piers are not set against a structure, reducing the possibility that they supported an arch, and perhaps increasing the possibility that they supported a tetrastylon of 4 free-standing columns without an overarching structure, as at Ephesus in the mid-6th c. Nothing is known about the upper parts of the structure.
The second phase of encroachment / decay is described by Hurst (1979) 44–46. We cannot easily distinguish ‘decay’ from the ‘demolition’ of the plaza in this sequence, although these may have been separated by a considerable length of time. This phase sees the construction of a number of buildings on the plaza, with earth and stone foundations, which may have had superstructures in mudbrick. They certainly had floors of beaten earth. These buildings, in the centre and on the edge of the plaza, add nothing to its monumental character. Their common character is suggested by their architecture and poor aesthetics in relation to what had previously existed within the forum. One structure with walls of compacted earth and small stones was even built over the back wall of the portico, which had been demolished. Furthermore, at least one burial is associated with these structures (building 8), whilst other burials have been found nearby; it has not been possible to date them firmly: Hurst (1994) 114–15. Of phasing, in the centre of the plaza, the foundations of these buildings overlay a cistern of 6th c. date, as mentioned above. The walls also occur after a sub-phase in which pits (Hurst (1979) 43–44) were dug and backfilled. The pits are not of themselves indicative of decay as the plaza covered a very large area, which although public, may have been more like a field or park than a traditional paved forum, and such pitting activity had been known earlier. Hurst does suggest that one building might have been associated with some of these pits, without explaining how. In a later report, the installation of a large pottery kiln is also mentioned, although not described: Hurst (1994) 114–15. Of dating, the backfill of the pits contains a very high concentration of ceramics, especially local amphorae, all of which “appears to be of late sixth-century date”, which provides a TPQ of ca. 587.5 for the second period of encroachment (based on the late 6th c. being equivalent to 587.5–600). Chronological indicators for building 8 and the burials consist of “only a few sherds dating to the seventh century”, alongside many residual pieces, which can be taken as associative ‘use’ finds. The well / cistern west of former Temple 2 produced a large assemblage of 7th c. pottery, mainly flagons, which provides associative dating of the occupation at this time. The well also contained a nasty surprise. Horse remains were mixed with rubble above the lowest fills, sealed by fills of pottery of this phase. This suggested to the excavators that there had been a deliberate defilement of the water supply, perhaps after the Islamic conquest, to prevent a Byzantine fleet establishing itself here: Hurst (1979) 44–46.
Overall, the second period of encroachment is probably best dated from 587.5 to 698. There is no evidence to suggest that the sequence goes beyond the 7th c., whilst the poisoning of the water supply might well have taken place on / shortly after the Islamic conquest of the city in 698 (e.g. Theophanes A.M. 6190 (AD 697/98 date in edn. of Mango and Scott (1997)). At this time, a decision was taken by the new rulers to abandon the city and to move the main regional centre to nearby Tunis. It would thus have been necessary to prevent any incoming ‘Byzantine’ fleet from holding the island.
Dating summary (first phase of encroachment / degradation): range 250–525, midpoint 387.5, class Cs5 (catch-all, aesthetic disjuncture), Cs9 (contextual ceramics), publication 3/3.
Dating summary (first phase of renewal): range 500–525, midpoint 512.5, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (second phase of renewal, comprehensive rebuilding, with new ?tetrastylon entrance): range 537.5–562.5, midpoint 550, class Cs9 (contextual ceramics), Cs5 (catch-all masonry), Cs3 (associative, phase of development), publication 3/3.
Dating summary (beginning of second phase of encroachment / decay): range 587.5–698, midpoint 642.75, class Cs7 (TPQ ceramics), Cs3 (associative, finds), x (historical text), publication 3/3.
07AFR Belalis Maior: There is no sign of 4th c. decay in this plaza, but rather it seems to begin in the 5th c. For a summary, see T.W. Potter, Towns in Late Antiquity: Iol Caesarea and its Context (Oxford 1995) 66–67.
Of repair, although one of the porticoes collapsed / was demolished, pieces of its inscribed frieze were incorporated in the paving of the north portico (Mahjoubi (1978) 174–75) and in a water channel of an adjacent street: see appendix C11 and K4a, with the water channel dated to after 317 and probably before 439, when the Vandal conquest seem to have put an end to even small-scale civic building works.
Of occupation, honorific bases on the forum stop during the time of Gordian III, although the traditional occupation level was maintained in the 4th c.: A. Mahjoubi, “La cité des Belalitani Maiores: exemple de permanence et de transformation de l’urbanisme antique”, AfrRom 1 (1984) 63–72; A. Mahjoubi, Recherches d’histoire et d’archéologie à Henchir El- Faouar. La cité des Belalitani Maiores (Tunis 1978) 137–75. An indication of occupation may come from the graffiti scratched on the forum and portico stylobate paving, which included crosses, as well as a circus drawing, a stylised garland and 6 cups arranged in a triangle (the latter on a step leading into the baths): Mahjoubi (1978) 142 with pl. 8a and p. 143 fig. 45a–b (of which the first looks like a cross-in-square gameboard) and p. 144 fig. 45c. The presence of inscribed crosses in public space is usually dated after AD 400: see dating foreword. This certainly suggests that any decay in the forum did not begin until after AD 400.
Of decay, the rooms opening onto the porticoes of the forum were subdivided in the 5th c., probably to serve as habitations. No reasoning is given by Thébert for the dating: Y. Thébert, “L’évolution urbaine dans les provinces orientales de l’Afrique romaine tardive”, Opus 2 (1983) (99–131) 111. The divisions can perhaps be seen on Mahjoubi (1978) 144 pl. 9a (east portico), 146 pl. 10a (north portico) and are alluded to on p. 175. Note that Thébert does not say here that the forum was reoccupied in the 5th c. by a series of private houses, as reported in A. Leone, Changing Townscapes in North Africa from Late Antiquity to the Arab Conquest (Bari 2007) 136, which is a misreading of the subdivision of rooms leading onto the north and east porticoes, described above. Mahjoubi (1978) 447–48 does not describe any other changes in the forum area, in resuming his concluding description of it.
Of further decay, the paving of the forum was covered by a layer of very friable dark earth some 15–20 cm thick, containing coins, especially of the end of the 4th c. to the Arab period, rims of ARS D and ‘late Christian lamps’, through which were cut some inhumation burials, which covered the whole of the forum: T.W. Potter Towns in Late Antiquity: Iol Caesarea and its Context (Oxford 1995) 66; Mahjoubi (1978) 173. There is nothing to date the burials, as far I could see, which could have been set here a long time after the earth layer formed.
Overall, it is especially significant that the bulk of coins in the dark earth layer do not date before the end of the 4th c. This suggests that cleaning stopped within ca. 25 years of 387.5, when one would expect such coins to be especially present in circulation, whilst the crosses on the paving suggest occupation persisting until at least AD 400. Thus, under the rules of this study, I would suggest a range of 400–412.5 for the beginning of formation of the dark earth.
Dating summary (beginning of decay): range 400–412.5, midpoint 406.25, class Cs8 (contextual coins), Cs1 (artistic style), publication 2/3.
07AFR Uchi Maius: Archaeological investigations in two separate zones of this forum suggest that the plaza experienced degradation in Late Antiquity. For a good summary (one of many), see S. Gelichi, M. Milanese and M. Biagini, “Uchi Maius—Henchir Ed Douamis: L’area del foro”, in Uomo, territorio, ambiente. La cooperazione italo-tunisina nel settore archeologico / Homme, territoire, environnement. La coopération tuniso-italienne dans le domaine archéologique, ed. A.M. Corda (Tunis-Cagliari-Sassari 2002) 34–37. See also M. Khanoussi and A. Mastino, “Nouvelles découvertes archéologiques et épigraphiques à Uchi Maius (Henchir ed-Douâmis, Tunisie)”, CRAI 144.4 (2000) (1267–1323) 1308–18. A full study is M. Biagini, La trasformazione delle aree forensi tra tardo antico e altomedioevo nelle città dell’Africa Proconsolare. Il foro di Uchi Maius (campagne 1995–2001) (Ph.D. Univ. of Siena 2002) (not seen). There were two excavations in the north-east zone, whilst a third took place in the south-east zone. Their sequences are complex and need to be considered separately, in terms of phasing and dating, before an overall picture of the plaza can be offered.
In the north-east zone, the first excavation revealed a cistern, built against an equestrian statue base for Septimius Severus. This cistern was set directly over the paving, before any silting took place. Against this cistern was built a series of rooms that extended over the north-east portico. These were set around a courtyard and made out of architectural elements, paving and honorific monuments taken from the plaza. Of dating, we can look at the last honorific dedication from the forum area, as providing a generic TPQ for the beginning of degradation implied by the cistern: this is a dedication to Valens, which was found out of context, built into a later structure. This is significant, as it is likely that a cistern would have only been built in the plaza when honorific dedications had ceased. For the inscription, see CIL 8.15452 = Lep. 235 = A. Ibba ed., Uchi Maius 2. Le iscrizioni (Sarda 2006) 173–75 no. 55. The Valens inscription is dated to AD 364–75, rather than the full range for this emperor of 364–78, because the death of Valentinian I resulted in two western emperors and the text implies that there was only one: S. Gelichi and M. Milanese, “Problems in the transition towards the Medieval in the Ifriqya: first results from the archaeological excavations at Uchi Maius (Teboursouk, Beja)”, Africa Romana 12 (1998) 457–86; C. Vismara, “Aetas succedit aetati. Il riuso di elementi classici nelle città africane fino all’età islamica”, in Atti X Giornata di archeologia. Il passato riproposto. Continuità e recupero dall’antichità ad oggi, ed. B.M. Giannattasio (Genoa 1999) (69–85) 73. The inscription provides us with a TPQ of 364. However, we also have some dated stratigraphy from the plaza itself.
In the north-east zone, the second excavation revealed a separate late building. This was located only a few metres away from the first excavation, in ‘area 2100’. The new late building was constructed against the other side of the base of Septimius Severus. Unlike the cistern it was erected after the forum had been covered by a uniform level of water-borne silt, on average ca. 14 cm thick, across the excavated area. A key report is S. Gelichi and M. Milanese, “Uchi Maius: la Cittadella e il Foro. Rapporto preliminare sulla campagna di scavo 1995”, in Uchi Maius 1. Scavi e ricerche epigrafiche in Tunisia, edd. M. Khanoussi and A. Mastino (Sassari 1997) 70–94, esp. 74–76, with plan on p. 77 fig. 30. Associated with this new building was a hearth. This occupation was followed by a further period of silting, of ca. 80 cm thick. Early in a second silt phase, another hearth was established against the wall of the same late building. Of dating, the first hearth has produced fragments of an amphora of type ‘Africana grande’. The amphora is given as dating to between the 4th and 5th c., although the specific type number is not provided (p. 74). The second hearth has produced a fragment of ARS D of Hayes 104A (p. 76), given as dating to the period between 500 and 580 [form 104A is now 487.5–650 in Bon with 5 sub-types, only one of which is before 525]. Taken together they technically suggest a contextual date for the hearth’s sub-phase of 487.5–512.5, based on the start date of the last find, the upper range of which provides a TAQ for the start of the silting. This is making an assumption that the hearths represent a connected sub-phase of activity rather than being entirely unconnected. There is no means of checking the subtype of Hayes 104A against Bon, as the report gives little detail. This issue and the lack of clarity on the amphorae means that the date is a poor one and may well be wrong once the 104A subtype is identified.
In the south-east zone an opus africanum building on the south-east corner of the forum (area 2200) was excavated, which was converted into an oil press in Late Antiquity: trapezoidal marks were found on a stone used here, which also appear on better-preserved press structures, elsewhere on site. Within the building, a floor was laid in reused slabs, including stones inscribed with texts dating from AD 180 and 196–97: Gelichi and Milanese (1997) 74–75. Further details were revealed by subsequent seasons of excavation / study, which reinforced this identification, as reported in M. Biagini and L. Gambaro, “L’area 2.200 (Foro): il frantoio”, in Uchi Maius 3. I frantoi, miscellanea, ed. C. Vismara (Sassari 2007) 195–215, esp. 195–204 with new plan on 195 fig. 6.1. This plan shows a second possible press, with a counterweight, cut from a reused statue block, and a series of basins, alongside other press features. There was also a store with amphorae set in the floor, adjacent to the structure, set inside the forum portico (room 7). Of dating, for the south-east zone there was initially no ceramic evidence from the press phase, only from the next phase, when a lime-kiln was installed over its ruins: Gelichi and Milanese (1997) 74–75. However, Biagini and Gambaro (2007) 204 give detail of further pottery from the foundation trenches of the building (“dalla fossa di fondazione del castello ai lati del contrappeso”) and from the floor of the store room. This pottery suggests to them that the press facility dated to the last quarter of the 5th c. or to the first years of the 6th c., with 50 years given for the operation of the facility (e.g. p. 205). Unfortunately, the ceramic report appended to Biagini and Gambaro (2007), which I use here as my dating basis, appears to refute this chronology.
The ‘appended ceramic report’ presents the following new dating for both phases in the south-east zone:
(i) Of the oil press period, from a floor surface (US 2.235, which should be the store room), come two late antique African amphorae, of unknown form. The first amphora is comparable with Keay 25 and Keay 20 (given as 4th to mid 5th c.) [Keay 25 = Africana 3 is 300–450 on RADR + Bon; Keay 20 is 300–450 in Keay 1984, although dating likely superseded], whilst the second amphora is comparable with Keay 36B (given as just before the mid-5th c. to the beginning of the 6th c., being the find with the latest start date in this layer) [now Keay 36 is 387.5–500 Bon]: Biagini and Gambaro (2007), 207–208. This suggests, under the rules of my study, a poor contextual date of 387.5–412.5 for the press, with the revised dating of Keay 36B being the last dated find, although of course the amphorae here are not confirmed identifications, only suggested ones, and might be imitations.
(ii) Of the subsequent lime-kiln period, from a ?floor layer (US 2.211, see map of Gelichi and Milanese (1997) 77 fig. 30) come a large number of sherds of 5th to 7th c. date, of which the latest wares are as follows. Of ARS D, there is 1 sherd comparable to Hayes 89 / Hayes 90A–B (given as from late 5th c. into the 6th c.) [Hayes 89 is 375–512.5 in LRP + Atlante; Hayes 90A is 540–560 in Bon and 90B is 587.5–612.5 in Bon], 6 sherds comparable to Hayes 104A (given as last decades of the 5th c. and first half of the 6th c.) [104A is now 487.5–650 in Bon, which dating I adopt], a sherd of El Mahrine 18 = Fulford 52.102 = Atlante I table 46.8–10 (given as second half of the 5th c. and first quarter of the 6th c. or ca. 450/ 60–520) [Atlante pp. 100–101 suggests it is Atlante table 46.9 or 46.10, I was not able to check Fulford or other references], 1 sherd similar to Hayes 80/81 = Fulford 24 (given as sometime in the 5th c.) [both 80 and 81 are 440–500 in LRP], and a lamp Atlante X = Hayes 2 (given as of 6th to 7th c., so being the find with the latest start date in this layer) [Atlante X is 425–700 in Bon, with many sub types, with Hayes 2 in LRP being ca. 425–550]. Other pottery, such as coarsewares, does not change this dating spread. Where dated, such coarsewares are predominantly of late 5th to 6th c. date (so giving them also the latest start date). Two other contexts in the lime-kiln phase produced two sherds, each of slightly later cooking wares: these were comparable or identical to wares considered to date to the 6th to first half of the 7th c. This list of ceramics is from Biagini and Gambaro (2007) 207–15, esp. 208–10. As the forms are given as ‘comparable to’ rather than as definite identifications, I will exercise deference to the dating given by the ceramicists for individual wares. Strangely, there is no mention of the ceramics from the foundation trenches, as described in the main report of Biagini and Gambaro (2007) 204, as if these have been reclassified. N.B. that I treat the finds in the floors (of both the press and the lime-kiln phase) as being construction fills, rather than as use-finds accidentally trampled into the floor.
Overall, the dating evidence for the decay of the forum is sparse and relates to different events, distinct in time, rather than being part of a single process. The dating can be summarised as follows: (i) The sequence of the north-east zone, along with the presence of the Valens inscription, indicates that the forum silted over sometime between 364 and 512.5, based on a TPQ from the Valens inscription and a TAQ derived from the upper end of a contextual date of 512.5 calculated for the hearth’s sub-phase. (ii) The sequence of the south-east zone testifies to the installation of an olive press, of ca. 387.5–412.5, a contextual date provided by the presence of the amphora comparable to Keay 36B, found in its floor, which is the find with the latest start date. (iii) The ruin of the press, with the installation of a lime-kiln, seems to date from the late 5th c./ early 6th c. This dating is based on the overall distribution of ceramics from the related floor and also on the start dates of the coarsewares and of the sherd comparable to Hayes 104A [500–580 ca. in Atlante]. This gives us a contextual date of 500–525 for the beginning of the lime-kiln phase. (iv) Minor contexts in the south-east zone of the forum suggest that occupation there continued through the 6th c., possibly into the 7th c.
Dating summary (for cistern, marking initial degradation, before silting): range 364–512.5, midpoint 438.25, class Cs7 (TPQ inscription), Cs3 (associative, phase of development), Cs9 (contextual ceramics), Cs8 (contextual coins), publication 2/3. Poor.
Dating summary (for silting in north-east zone): range 364–512.5, midpoint 438.25, class Cs7 (TPQ inscription), Cs3 (associative, phase of development), Cs9 (contextual ceramics), Cs8 (contextual coins), publication 2/3. Poor.
Dating summary (for installation of rooms with olive press in south-east zone, after phase of silting): range 387.5–412.5, midpoint 400, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (for beginning of lime-kiln phase after phase of ruin): range 500–525, midpoint 512.5, class Cs9 (contextual ceramics), publication 3/3.
07AFR Lepcis Magna (Old Forum): A rising of the forum surface level seems to have occurred, probably after the last late antique repairs to civic monuments and certainly before the Byzantine reoccupation. A gate in the Byzantine fortification shows this higher level at a gate leading into the cardo (+0.5 m), as do higher occupation features on the plaza. A level of 0.87 m higher than the Early Imperial level is recorded near the Severan exedra on the old forum [which is just in front / north-west of the basilica by ca. 13 m], with a cistern, a well and a Christian tomb at this height: M. Pentricci, “L’attività edilizia a Lepcis Magna tra l’età tetrarchica e il V secolo: una messa a punto”, in Leptis Magna. Una città e le sue inscrizioni in epoca tardoromana, edd. I. Tantillo and F. Bigi (Cassino 2010) (97–171) 141–42 with references. A TPQ for the abandonment of the ancient surface is probably the date of the last honorific statue base to be recovered from the forum vetus area (reused in church [on west side of forum], built into the apse): marble base for a statue of PLRE 1.632 Flavius Archontius Nilus 1, comes and praeses (governor) of Tripolitania, patron of Lepcis, by Leptimagnenses. IRT 563 = LSA 2185. Dated to AD 355–361, based on Nilus’ career. This probably comes from a display setting in the forum and would unlikely have been set within a plaza full of silt: it implies care for the ornament and repair of traditional civic spaces, of which the most basic act would be cleaning away silt from the surface of the forum. A TAQ for the raising of the forum level seems to be the construction of the city wall, dated to 533 to 543 based on texts (see appendix V4c). It is important to point out that the forum paving does not seem to have been spoliated: L. Lavan site observations, based on a tourist photo of http://henrysuter.ch/photos/Tourisme/2008/Libye-nord/slides/dsc002660.html (last accessed June 2018).
Dating summary (beginning of decay): range 355–543, midpoint 449, class Cs7 (TPQ inscription), x (historical text), publication 3/3.
07AFR Lepcis Magna (Severan Forum): An almost total stripping of the paving of the open area of the forum occurred, except in the porticoes, where the paving was finer. There was then a build-up of silt, probably after the last late antique repairs to civic monuments and certainly before the Byzantine reoccupation, as the area had been covered by 0.5 m of flood sand by the latter time: J.B. Ward-Perkins, “Excavations in the Severan Basilica at Lepcis Magna, 1951”, PBSR 20 (1952) (111–21) 111 (not seen). This reference was derived from M. Pentricci, “L’attività edilizia a Lepcis Magna tra l’età tetrarchica e il V secolo: una messa a punto”, in Leptis Magna. Una città e le sue inscrizioni in epoca tardoromana, edd. I. Tantillo and F. Bigi (Cassino 2010) (97–171) 152. A TPQ for the beginning of decay in the forum can be derived from the last honorific statue base known from the forum. This is (from in front of the east portico) a reused base for a statue of PLRE 2.813 Fl. Ortygius, comes et dux of Tripolitania by ordo and populus: Tantillo / Bigi no. 31 = LSA 2177. Dated to AD 408–423, based on years of joint reign of Augusti Honorius and Theodosius II. A TAQ for the raising of the forum level seems to be the construction of the city wall, dated to 533 to 543 based on texts (see appendix V4c).
Dating summary (beginning of decay): range 408–543, midpoint 475.5, class Cs7 (TPQ inscription), x (historical text), publication 3/3.
See also (appendix K4a): 06ITS Rome (Forum of Caesar): fifth late antique phase, one of decay, saw the stripping of the paving in two tabernae (nos.10 and 11) that have been recently excavated on the south-west side of the forum, in the period 400–425; 06ITS Ostia (Palaestra of the Forum Baths): In phase 7 rooms were built inside the portico, at a higher occupation level than the ancient plaza. This can be treated as ‘decay’, although the revetment had already been stripped off in an earlier phase. This probably happened in the period 443–?450.
See also (appendix V1): 06ITS Herdonia: An abandonment began sometime in the period 379–500.
See also (appendix X1d): 02HIS Tarragona (Tarraco) (Lower (Civic) Forum): The destruction of the civil basilica by fire occurred sometime AD 361–86 based on a hoard trapped under a collapsed column.
Discounted: 02HIS Valencia (Valentia): Some publications state that the forum became disused in the 5th c., with a cemetery established and a small chapel built over it, and that a cathedral was then built next to the chapel in the 6th c. However, this sequence of events relates to the area south-west of the forum, not the forum itself, of which only a small part has been uncovered. It involves a baptistery, mausoleum and cemetery. The cemetery seems to be dating from 440–560. See appendix A1 for dating and also map in A.V. Ribera i Lacomba, “Origen i desenvolupament del nucli episcopal de València”, in VI Reunió d’arqueologia cristiana Hispànica. Les ciutats tardoantigues d’Hispania: cristianizació i topografia (València, 8 al 10 de mayo de 2003), edd. J.M. Gurt and A.V. Ribera i Lacomba (Barcelona 2005) (207–43) 215 fig. 8. I have not been able to access the final excavation reports of this site (see appendix A1), which are stored in a Valencian archive, which might include other evidence of the disuse of the forum, not known to me.
V4c Degradation and Disuse in the West, 6th C.
02HIS Conimbriga: The destruction, spoliation and reoccupation of the forum is indicated by stratigraphy that overlies the buildings surrounding it. However, I have not been able to obtain details about exactly which buildings were affected, in what order. This stratigraphy is dated on the basis of the presence of ARS D Hayes 103, Hayes 104a, Hayes 108, Hayes 110 and Late Roman C, coming from related spoliation and occupation levels: J.M. Gurt Esparraguera, “Transformaciones en el tejido de las ciudades hispanas durante la Antigüedad tardía: dinámicas urbanas”, Zephyrus 53–54 (2000–2001) (443–71) 452. These ceramics are respectively given dates of 500–75 [LRP], 487.5–650 [Bon, with 3 subtypes], 600–612.5 [LRP], 450–512.5 [SLRP], with the unspecified Late Roman C ranging from 375 to 650 [LRP]. He refers to his own article J.M. Gurt Esparraguera, “Topografìa cristiana de la Lusitania. Testimonios arqueológicos”, in Los últimos romanos en Lusitania, edd. A. Velazquez, E. Cerrillo and P. Mateos (Cuadernos Emeritenses 10) (Merida 1995) (73–95) 88–89. In this second publication, he specifies that the wares listed above are from the Insula del vaso fálico (south-west of the forum). He also adds to the list Hayes 97 [400–550 in LRP], noting that the forum destruction coincides with the presence of Hayes 99 [487.5–612.5 in Bon]. He also refers to the following report: J. Alarcão and R. Étienne, Fouilles de Conimbriga I: l’architecture (Paris 1977) 240, which deals with the destruction level of the Insula del vaso fálico and lists these ceramics. However, this old report does not list the ceramics by context (as far as I could see) and it inevitably contains some older dates given for them (not drawing on SLRP of 1980 for example). Thus, Gurt Esparraguera (1995) and (2000–2001) are to be preferred to it as a basis for the ceramics. The destruction of the forum and this insula seems to be one and the same phase, as far as I understood from Gurt Esparraguera (1995) 88–89. Gurt Esparraguera (2000–2001) 452, opts for a date range of the whole of the 6th c. for the destruction. He clearly rejects an earlier date for the destruction of the forum, once believed to have been the result of an attack by the Suevi sometime in AD 465–68: Gurt Esparraguera (2000–2001) 452. Overall, I can see no ceramics listed by him with start dates later than AD 500, based on the dates provided in LRP/SLRP and Bon, so I adopt the range 500–525 for a contextual ceramics date, taking the Hayes 103 as the last find.
Dating summary (destruction): range 500–525, midpoint 512.5, class Cs9 (contextual ceramics), publication 3/3.
02HIS Carteia (near Cádiz): The stairs of the Forum temple, which led down onto plaza, were blocked by a new building, which the excavators place in the 4th–5th c. After this, a cemetery was established over the forum, which the excavators place in the 6th c.: D. Bernal Casasola, “Carteia en la antigüedad tardía”, in Estudio histórico-arqueológico de la ciudad de Carteia (San Roque, Cádiz) 1994–1999 , edd. L. Roldán Gómez, M. Bendala Galán, J. Blánquez Pérez, and S. Martínez Lillo, vol. 1 (417–64) 453–55, with plates 1–2 for maps of the forum.
Of the new building, which blocks the temple stairs, the documentation provided by the report is poor. We are told to see section II.2.7.1 on pp. 235–47, but this does not have details, except in a table on p. 246, where a wall (F.T./18) is described, without finds. We are also told that two tombs belong to this phase. However, they were excavated by an early team, which does not seem to have produced a report worth quoting. Rather, the chronology of the new building phase now depends only on the recorded fill of two associated pits (F.T./15 and F.T./13), one of which is a spoliation pit relating to the adjacent temple, and on related redeposited strata: (i) The first pit (F.T./15) has produced ARS given as dating to “VII; f III–IV d.C.”, alongside amphorae and coarsewares, for which no date is given. (ii) The second pit (F.T./13) has produced ARS and lamps [given as 4th–6th c.], alongside amphorae and coarsewares given as 6th c. A second fill within this second pit produced ARS D and a coarseware casserole attested at Carthage (in contexts dating between 475/500 and 575/600). We are told that recent studies date this casserole, which comes from a volcanic area of the western Mediterranean islands, from the beginning of the 5th c. through all of the 6th c. (J.M. Macías Solé (1999) La cerámica comuna tardoantiga a Tàrraco. Análisi tipològica i històrica (segles V–VII) (Monografies Tarraconenses 1) (Tulcis 1999) (Tarragona 1999) 63–64, form Ca/Lip8, lám. 6, 8.7 (not seen)). (iii) From adjacent redeposited strata relating to the second pit (F.T./5, discussed on pp. 454–55) came finds of an African lamp of Atlante X (their figure 284 no. 2, which could be Atlante fig. 101.1 (lamp X.B2), although the match is not a very close one and the fragment is small; not dated in Atlante, although B1a is given as late 4th to possibly 6th c. on p. 200) and ARS, including Hayes 99 (given as being ca. 500, although 487.5–612.5 in Bon, so perhaps meaning beginning in AD 500), plus other coarseware. The excavators believe that the material indicates a date of the first half of the 6th c. In particular, they note the appearance in this context of amphorae Almagro 51c and Keay 16, the production of which they believe lasts until the beginning of the 6th c. [find numbers confirm they mean ‘this context’ not ‘these contexts’, as in the main text] [Almagro 51c is 200–?450 RADR; Keay 16 is 187.5–450 ICAC].
Of the subsequent cemetery, established across the whole forum, we know little. Although it is referred to often in section II.2.7.1 (e.g. 455), it does not appear to have been studied by the present excavation team. It is referred to as being of later 6th c. date (p. 453). However, inhumation clearly began at the time of the previous phase (two tombs mentioned above). Until evidence is presented that contradicts this, I will place the start of the cemetery in the previous phase with the same dating.
Overall, the beginning of decay in the forum, as manifested by pitting, should be given a fairly weak [i.e. disturbed] contextual date of the first quarter of the 6th c., based on the ensemble of pottery related to the second pit (for which the record is clearest). This is because the latest ceramics from the fill of the pits date from ca. 500 onwards, in their calculations. I must myself modify this to begin in 487.5, based on the start date of the last ceramic according to Bon, under the rules of this study. Although, one might want to place the building over the temple steps before this phase. I can see no stratigraphic basis to do so.
Dating summary (for decay of forum, including beginning of cemetery): range 487.5–512.5, midpoint 500, class Cs9 (contextual ceramics), publication 3/3.
06ITS Luni: The buildings around the forum were redeveloped, with demolition and levelling being detected over the civil basilica, within a city that otherwise saw a good deal of late antique occupation: B. Ward-Perkins, “L’abbandono degli edifici publici a Luni”, Quaderni del Centro di Studi Lunense 3 (1978) 33–46. This occurred in the 4th–5th c., allegedly, although I have seen no evidence to support this dating. After the denudation of the paving, ca. 30 cm of silt accumulated, after which the site was reoccupied. Houses were built, which were detected in a quadrant of the forum that was carefully excavated: B. Ward-Perkins, “Two Byzantine houses at Luni”, BSR 49 (1981) (91–98) 91–92. The deposit under house 1 contained a large quantity of pottery “which seems on first examination to be of the later sixth century”. A coin of Justin II (AD 565–78) was recovered from the reflooring of house 2, and a radiocarbon date of AD 640 ± 80 years was obtained from an oven in house 3. The best basis for dating the decay of the forum seems to be the pottery under house 1. B. Ward-Perkins (pers. comm. Jan 2016) agreed with me that this ceramic deposit was likely a primary rubbish dump, rather than an imported fill, given the nature of the building. However, he also noted that the dump was found on top of the ca. 30 cm thick layer of (sterile) silt that lay over the robbed surface of the forum. This could suggest an abandonment of the forum in perhaps the 50 years prior to the rubbish dump, although this is a rough estimate, with no TPQ. It depends on rough estimation of time needed for a phase of development, longer than 25 years (my usual phase length), as the silt is deep. Quite when the first degradation of the plaza took place, with the demolition of buildings and the robbing of paving, is not certain. The plaza may have remained in use in a degraded state, with paving robbed etc., without being abandoned, as seems to be the case at fora elsewhere in the west in the 5th c., notably at Nora.
Dating summary (for abandonment of the forum): range 500–50, midpoint 525, class Cs9 (contextual ceramics), Cs3 (associative, phase of development), publication 2/3 (as ceramics not specified). Poor, as no TPQ established.
06ITS Rome (Temple of Peace / Forum of Vespasian): In the south area of the plaza, which has been excavated recently, podia and euripus structures were demolished and the site was covered with a new set of buildings. These were arranged around a large open area, paved with bipedales (bricks), surrounding a basin lined with marble. Two further basins, communicating with each other via ceramic tubes, were also found within the structure. The buildings were probably single-storey in height, on account of their very shallow foundations. They were made with much reused material. One that survives well was paved in bipedales (bricks). Dimensions are not available, partly because the edges of the buildings have not been reached by the excavation. The south-west portico of the plaza was paved in a floor of cocciopesto and given rooms built in opus vittatum. These developments have been dated to the very first years of the 4th c., based on brickstamps. I will take this as meaning the years 300–312.5 or ‘early 4th c.’, under the rules of this study. The buildings are interpreted as storage or market buildings. It is supposed that their construction was necessitated by the transfer of activities of this nature away from the Basilica of Maxentius area, in order to build that structure. Subsequent repairs and modifications are known, such as the substitution of the paving with marble, but the works do not seem to constitute a change of function of the complex: R. Santangeli Valenzani, “I fori imperiali nel medioevo e nell’età moderna”, I fori imperiali: Gli scavi del comune di Roma (1991–2007), edd. R. Meneghini and R. Santangeli Valenzani (Rome 2007) (114–22) 115–17 with p. 116 fig. 121. I am not myself currently prepared to accept this site as representing a market building, based on what is known of the plan, its fixtures or the finds.
A cemetery developed in the 6th c., when the plaza was levelled (or a short time after this time): see the preliminary report, without dating details, of R. Santangeli Valenzani, “I fori imperiali nel Medioevo”, RM 108 (2001) (269–83) 269–71, with p. 270 fig. 1 (plan). See also a popular summary in Santangeli Valenzani (2007) 119, specifying that the cemetery was in the southern part of the plaza, with plan on fig. 124.
Note that Procopius saw a fountain standing before the ‘agora which the Romans call Forum of Peace’ during his time in Italy in 537–40 (Procop. Goth. 4.21.12), and that he attested to the survival of bronze statues by Pheidias and Myron, including a calf and a bull (Procop. Goth. 4.21.11–14). As in the case of Arles, we need not think that the archaeology contradicts the literary evidence, as the plaza was very large, and could have developed in a number of different ways, or have referred to slightly different moments in its 6th c. history.
Dating summary (‘market building’): range 300–312.5, mid-point 306.25, class Cs6 (absolute, inscription, in situ), publication 2/3.
Dating summary (cemetery): range 500–600, midpoint 550, class 0, publication 1/3. Pending.
06ITS Rome (Forum Romanum): A sondage inside the East Rostra revealed some important stratigraphy relating to the decay of at least parts of the plaza: C.F. Giuliani and P. Verduchi, L’ area centrale del Foro romano (Florence 1987) 165–67. This concerns two layers, (2) and (4), that are stratigraphically below layer (1), which was cut by a robbing trench. Layer 1 had Roman and Early Medieval ceramics (vetrina pesante [sometimes called ‘Forum Ware’], but no Lazio pottery) and modern material. But layers 2 and 4 were uncontaminated and seemed to relate to artisanal activity involved in the recycling of building materials. Layer 2 included cinders, small pieces of charcoal (frustoli di carbone), flakes of different kinds of marble, animal bones and ceramics, as detailed below. Layer 3, below 2 and above 4, was a compact layer of clayey earth, with diverse building materials, but no useful dating evidence. Layer 4 was similar to 2, with cinders and marble flakes plus slag and pieces of lead clamps, alongside a small amount of ceramics. Under these layers, a fill with bricks and mortar fragments was found, which seemed to relate to the preparation of a floor surface, after the monument had lost its primary function: Giuliani and Verduchi (1987) 165.
Of dating, layer 2 contained ARS D Hayes 99 [487.5–612.5 in Bon] and Hayes 103 [500–575 in LRP], vetrina pesante, red-painted and ceramica acroma, including little comb-decorated amphorae (a pettine) [perhaps Keay 62 based on http://badwila.net/amphorae/keay_LXII/index.html (last accessed Sept 2017), which is split into various subtypes between 466 and 650 in Bon], and open vessels that imitate ARS, as well as some residual pottery. Layer 4 contained ARS Hayes 104 [500–625 in LRP, now 487.5–650 in Bon], lamps of type Provost 10B [not confirmed by me], vetrina pesante, amphorae a pareti costolate [not confirmed by me] and other finds suggesting probably the 6th c. AD: Giuliani and Verduchi (1987) 165. I am not able to check all of these wares, but the ARS certainly suggests a date in the 6th c. As I do not yet know what the start date of the latest ceramics is from the group [Hayes 103 with 500–700 is the latest confirmed], I will defer to the date provided by the ceramic report and adopt the whole 6th c., although under the rules of this study I should normally prefer a date in the first quarter of the 6th c. I must also note that vetrina pesante is normally placed from the end of the 8th c. onwards—http://archeologiamedievale.unisi.it/SitoCNR/Ceramica/AM/AM39.html (last accessed February 2017). Thus, there is substantial doubt over the correct dating. The absence of later forms of ARS suggest, that the 6th c. date is possible for this artisanal activity [I thank E. Vaccaro for this observation]. Thus, we must envisage either ceramic error in identifying the vetrina pesante, or that there was some undocumented disturbance of the site.
Away from the area of the sondage, it is clear that this late artisanal establishment occupied the whole area around two of the ‘two plinths’ [first two robbed out honorific column plinths from the north], and that rooms were connected by cutting a new passage. Furthermore, round pit holes were cut into the structure, prior to the plinths being robbed: the report writers thought these pits might have been used for fusing lead. In the area to the west of the East Rostra, a number of further structures were found by the earlier clearance excavations by P. Rosa, without dating evidence, indicating a marble workshop. The evidence included a large marble block that was being cut into veneers, alongside traces of wooden structures and fixtures for the raising up of loads [to allow loading onto vehicles]. These features indicate that the productive area extended onto the paving which was joined to the former East Rostra by a new threshold set into the perimeter wall: Giuliani and Verduchi (1987) 165. These workshop traces were well preserved on the forum paving, because they were covered with silt in a flood event that caused the area to be abandoned: p. 165 with p. 145 (for further references). It is nonetheless interesting that there does not seem to have been a spoliation of the forum pavement. Rather, the workshop was involved in the systematic stripping of selected monuments, whilst the basic functional capacity of the plaza was preserved. Nonetheless, the appearance of a site of primary production within the plaza, rather than small-scale artisanal production for retail, marks a radical change.
Dating summary (beginning of decay): range 500–600, midpoint 550, class Cs9 (contextual ceramics), publication 3/3. Uncertain.
07AFR Bararus (Rougga): The temples were demolished, the colonnades knocked over, and the paving stripped to its substratum almost everywhere. The area was covered over with a layer of earth mixed with gravel. Only the column bases remained in situ. A ‘Byzantine’ dwelling was established ca. 0.5 m above the paving, along with a layer of sandy earth blackened by charcoal, some tens of centimetres thick. One wall of the new structures, built of dry stone walling, with fragments of reused architectural elements, cut through the earth and reached the stylobate of the north portico of the forum, which it followed. A series of walls was built at 90 degrees from this wall into the forum, creating a series of rooms, which are only known in fragments. Inside one of these rooms, resting on the floor level of the structure, was a coin hoard: R. Guéry, C. Morrison and H. Slim, Rougga III, le trésor de monnaies d’or byzantine (CÉFR 60) (Rome 1982) 11–14 with fig. 2 (plan) and fig. 3 (photo), showing the structures. This hoard produced 268 coins of the period from Maurice (582–602) to Constans II (641–668), of which the last were 14 coins of the latter emperor, with the year 647 being taken as the year of burial, from a comparison to other hoards: C. Morrisson, “Etude numismatique” pp. 59–74 (same volume), which takes historic significance in H. Slim, “Le trésor de Dougga et l’expédition Musulmane de 647 en Ifrikiya”, pp. 75–94. Although this is not very useful information in calculating a contextual date for the buildings, one might tentatively suggest a poor date of 668 for the TAQ of the occupation. Of course, what is needed is ceramics and coins from stratigraphic layers. I do not think we can assume the buildings are ‘Byzantine’. Other articles have dated the structures built around the forum entrance to the mid-6th c.
In another report, the last indication of the use of the drains is stated as being 3rd c. The stripping of the paving of the forum is described as dating to the second half of the 5th c. and first half of the 6th c. The installation of 2 cellars with amphorae against the south stylobate is given the same date, although without providing dating justification in either case. A subsequent layer (12) contained ceramics, which J.M. Hayes suggested dated to the three last quarters of the 6th c., whilst a subsequent layer (11) of ‘the end of the 6th c.’ [no justification given] corresponds with the installation of houses resting on the remains of the south portico and a building set over temple A, which might be a little fortress. Both were occupied in the first half of the 7th c. The end of occupation is dated by the coin hoard described above: R. Guéry, “L’occupation de Rougga (Bararus) d’après la stratigraphie du forum”, BCTH 17B (1981) (91–100) 97–98 (drains on p. 97, the other matters on p. 98). Despite the publication of a stratigraphic section and the existence of clearly labelled layers (e.g. p. 95 fig. 2), there are some problems with this report.
Overall, the report’s failure to cite specific finds to justify the dating provided is a major weakness, even if the recorded presence of John Hayes on site gives its dating some potential credibility. The pottery of layer 11 at least provides some indication of a TPQ for the houses built in the subsequent phase, and potentially a contextual TAQ for the start of decay. However, the absence of clearly documented finds to date the stripping of the paving, or to date the 4th c. occupation, is problematic. We have only a statement that the last material recovered from the forum drains was 3rd c. in date. This gives the beginning of decay only a TPQ of 200, with a TAQ of 575 derived from the top end of the date range of ceramics found in layer 12. Only fuller publication, with the finds, can produce a tighter dating at present.
Dating summary (start of decay of forum): range 200–575, midpoint 387.5, class Cs8x (contextual unknown finds), Cs9 (contextual ceramics), publication 2/3. Poor.
07AFR Sabratha: A cemetery over the forum has been dated to the Byzantine period, based on funerary inscriptions, a grave containing a coin of Justinian or Maurice Tiberius, and also pottery: P.M. Kenrick, Excavations at Sabratha 1948–1951 (JRS monograph no. 2) (London 1986) 34, 83–87. Of the funerary inscriptions, see R. Bartoccini, “Le iscrizioni sepolcrali nella basilica cimiteriale del Foro di Sabratha (Tripolitania)”, RACrist 51 (1975) 143–67 (with map of cemetery on p. 144), noting that the inscription on p. 145 no. 1 (found behind the western apse) contains the name Ferrandus, also attested at Carthage in AD 520–47 and that pp. 156–57 no. 16 (from a cemetery in the street area between the church and the Antonine Temple) contains the name George, a name which apparently comes to Africa with the Byzantine conquest; other Greek and Vandal-style names are present as listed on p. 162; N. Duval, Recherches archéologiques à Haïdra, vol. 1. Les inscriptions chrétiennes (CEFR 18) (Rome 1975) 484 with n. 6. Duval notes that the use of an indiction is clear evidence of a Byzantine date in Africa, as there is no example of an inscription using the dating formula of a Vandal king that uses this system. The indiction dating is used for the following epitaphs: e.g. Bartoccini (1986) 150–51 no. 9 and no. 10. Of the pottery, Kenrick (1986) notes (p. 34) that the backfill of two graves included, as their latest sherds, “rather more of the latest African forms of the late sixth and seventh centuries (Hayes forms 91D, 105 and 106).” These forms are dated, respectively, to AD 600–650 [LRP], 587.5–700 [Bon], and 600–660+ [LRP]. The connection of the two types of evidence is reassuring, as the excavator expressed doubt as to the stratigraphic sequence here (p. 3 n. 24).
Overall, the ceramics, assessed contextually, show that the cemetery was active until around 600, as that is the start date of the latest ceramic, although not necessarily later. In terms of a TPQ for the cemetery, it is reasonable to see the rebuilding of this centrally-located church as an ‘act of restoration’, appropriate as part of the reconquest of 533 (and the years that followed), which subsequently attracted the cemetery with its ‘Byzantine period’ epitaphs.
Dating summary (beginning of cemetery): range 533–600, midpoint 566.5, class z (regional development), Cs9 (contextual ceramics), Cs8 (contextual coins), Cs6 (inscriptions in situ), publication 3/3.
07AFR Lepcis Magna (Severan Forum): The forum was likely redeveloped, as its entrances were blocked up in the reconquest period, as part of its incorporation within the fortification circuit, and it was cut off from the judicial basilica, which was converted into a church. Buildings were established within the forum plaza, but were apparently removed without record by the Italian excavators: D. Pringle, The Defence of Byzantine Africa from Justinian to the Arab Conquest: an Account of the Military History and Archaeology of the African Provinces in the Sixth and Seventh Centuries, vol. 1 (BAR-IS 99) (Oxford 1981) 208ff. On the church see appendix U2. For blocking walls in the portico north of the Severan temple inside the forum, which probably date from this period, see: M. Pentricci, “L’attività edilizia a Lepcis Magna tra l’età tetrarchica e il V secolo: una messa a punto”, in Leptis Magna. Una città e le sue inscrizioni in epoca tardoromana, edd. I. Tantillo and F. Bigi (Cassino 2010) (97–171) 151 with fig. 4.39.
Of dating, Pringle (1981) 211 considers that the two phases of fortification, similar in style, which this blocking of the forum relates to, date from 533 to 543. In terms of a TPQ, the fortification should probably date after the designation of the city as seat of the Dux in 534: Cod. Iust. 1.27.2.1a (AD 534). In terms of a TAQ, the city was fortified by 543, as seen in Procopius’ account of an occasion when the leaders of the Leuthae were invited inside the city and slain at a banquet held by the Dux Tripolitianae: Sergius: Procop. Vand. 4.21.2–15.
Dating summary: range 534–43, midpoint 538.5, class x (historical texts), Cs3 (associative, phase of development), publication 3/3.
See also (appendix Z2): 07AFR Iol Casearea: The forum began to degrade with the destruction of a church by a fire, sometime ca. 490–515, the debris of which was not cleared away, accompanied by the demolition of the civil basilica. The phase was followed by the gradual accumulation of hillwash, into which some industrial activity was set. This industrial activity began in the period 500–600.
V4d Degradation and Disuse in the West, 7th C.
See also (appendix V4b): 07AFR Carthage: The open plaza on Admiralty Island (in the centre of the circular harbour) saw substantial encroachment from the 7th c., with the construction of earth and stone foundations for a number of buildings, which had earth floors. One of these buildings was built over the back wall of the portico, which had been demolished. At least one burial is associated with these structures, whilst other burials have been found nearby. A large pottery kiln was also installed. This phase is dated to 587.5–698.
V4e Degradation and Disuse in the West, Undated within Late Antiquity
06ITA Forum Traiani (Sardegna): Forum plaza encroached by late buildings with very mixed up reused building material, but no obvious architectural pieces: L. Lavan site observations, April 2016.
06ITA Paestum: In the south-west corner of the Forum is an encroaching building, constructed out of reused statue bases with a brick floor, established on one corner of the forum, part on the paving, ?part in the porticoes. One reused base has strange cuttings in it, perhaps a support for some sort of water heating for a bar? L. Lavan site observations, April 2016.
07PAN Nesactium: Paving was apparently removed from the forum for building works on the fortifications in the 5th–6th c.: R. Matijašić, “Foro e campidoglio di Nesactium (Nesazio)”, in ‘Forum et basilica’, in Aquileia e nella cisalpina romana, ed. M. Mirabella Roberti (Antichità Altoadriatiche 42) (Udine 1995) (121–39) 132. No dating evidence is provided to support this assertion.
Dating summary: Undated.
V5a Degradation and Disuse in the East, 3rd–4th C.
10MAC Butrint (Buthrotum): The forum paving is sealed by a rubble deposit that contains much marble veneer and many architectural fragments, suggesting degradation or destruction. Inside the deposit, which likely reflects the spoliation of the forum, the latest pottery is 4th c. Above this was a compact rubble deposit with mixed building debris, which the excavators believe is a levelling deposit, in which the latest (unspecified) pottery dates from the mid- to late 4th c.: D.R. Hernandez and Dh. Çondi, “The Roman forum at Butrint (Epirus) and its development from Hellenistic to Mediaeval times”, JRA 21 (2008) (275–92) 289. This suggests to me a contextual date for the initial abandonment of the forum, of sometime during the 4th c., taking the full range given for the unspecified pottery of the lower deposit.
Dating summary: range 300–400, midpoint 350, class Cs9 (contextual ceramics), publication 2/3.
10MAC Argos: The date of the monumental disuse of the agora is complex: some repairs to public buildings seem to co-exist with evidence of abandonment or decay, much of which is 5th c. The restoration of monuments around the agora can be seen in the repaired fountain, a new temple and a new arch, carried out in reused material, as noted in appendices K7b, Q1 and F7a. The arch can be dated to 350–400, based on unspecified ceramics, according to the excavators.
Evidence of decay on the agora can be seen in the installation of houses with 5th–6th c. occupation, in the western part of the Roman south portico (called the ‘extension du portique’), reported in P. Aupert et al., “Rapports sur les travaux de l’École française en Grèce 1977: Argos”, BCH 102 (1978) (771–802) 777 (based on unspecified coins and pottery); M. Piérart et al., “Rapports sur les travaux de l’École française en 1980: Argos”, BCH (1981) (891–917) 904 (no reason given, but stating that domestic occupation is after the end of the 4th c. and ends around the end of the 6th c.). A well and a foundry were also excavated, towards the centre of the Roman south portico. They occur within an occupation level of the plaza, which is dated by the last phases (sols II and I) of the ‘Square Monument’: P. Aupert et al., “Rapports sur les travaux de l’École française en Grèce en 1976: Argos”, BCH 101 (1977) (667–83) 673. These last phases of the square monument are given as dating to the 5th–6th c. in the report. However, sol II is believed to date to the end of the 4th c., based on unspecified coin finds, whereas layers immediately above sol II, subsequently covered by sol I, contain 5th–6th c. coins and ceramics, as spelled out in the earlier report of F. Croissant et al., “Chronique des fouilles et découvertes archéologiques en Grèce en 1974”, BCH 99 (1975) (696–708) 704–705. The earth of sol II contains fragments of the adjacent south portico (capitals, edging tiles, brick and masonry debris) (1975) 705. This debris extends to the bottom of sol II: sol III, beneath it, is covered by further collapsed architectural parts of the colonnade, such as capital fragments, brick masonry and the remains of 4 arcades: Aupert et al. (1977) 673–675 with 674 fig. 8. It has more recently been suggested that this sol II earth was part of a wider destruction across this part of the Agora: an ashy destruction layer near the dromos of the agora has 3rd–5th c. ceramics, with its last ceramics being of the 4th–5th c., and an amphora of the 4th–5th c.: M. Piérart, “Rapport sur les travaux de l’École française d’Athènes en 1990: Argos”, BCH 115 (1991) (667–86) 672 with 674 fig. 7. The filling of wells on the agora in the 5th c. AD need not be as significant. See a summary of references, not sufficiently examined, in A. Oikonomou-Laniado, Argos Paléochrétienne. Contribution à l’étude du Péloponnèse Byzantin (BAR-IS 1173) (Oxford 2003) 5–7.
Another important piece of evidence comes from the building of a substantial structure, measuring some 13.3 m by perhaps 30 m, with a raised floor in the west corner of the agora. This perhaps indicates a granary, which was built over and replaced the ‘Salle hypostyle’. The latter structure had been destroyed, perhaps at the same time as the south portico. Coins from robber trenches, very close in time, suggest that the Salle hypostyle was destroyed in the second half of the 4th c.: P. Aupert et al., “Rapports sur les travaux de l’École française en 1975: Argos”, BCH 100 (1976) (747–58) 754. From inside a wall of the ‘granary’ come 7 coins of the second half of the 4th c., including one of Valentinian I of 364–75 and one of Valentinian II of 378–83, along with fragments of lamps. These finds support a contextual date of 378–403 for the building of the ‘granary’: P. Courbin, “Chronique des fouilles en 1952: Argos”, BCH 77 (1953) 256. The floors within the ‘granary’ continue into the 6th c.: Courbin (1953) (243–63) 244. This sequence is synthesized by J.-F. Bommelaer and J. des Courtils, La Salle hypostyle d’Argos (Études Péloponnésiennes 10) (Paris 1994) 49–54. Based on this evidence, the destruction across the agora should be placed before ca. 403, under the rules of this study.
Overall, the absence of 5th c. materials from the destruction layers and their presence in the new occupation layers does suggest a disaster dating to around 400, although the site publications are far from adequate, especially as regards the identification of ceramics. The best evidence for a TPQ for the destruction comes from the unspecified coins of sol II, which suggested a date of the end of the 4th c. (for 387.5–400 under the rules of this study). This is supported by the unspecified ceramic and coin finds of sol I above. However, the best evidence for a TAQ comes from the granary. The destruction dates before the ‘granary’ was built over the Salle hypostyle. This ‘granary’ should, under the rules of this study, date in the 25 year period after 378, the start date of its last coin, which gives us a range of 378–403. The end of this date range (AD 403) can also serve as a TAQ for the preceding destruction of the agora in its traditional form. This TAQ is supported by the lack of finds of exclusively 5th c. date from the pertinent layers from the agora itself (i.e. sol II).
This is not the end of the agora, as the square monument saw two developments to coincide with sol II and sol I. In the first instance the edge of the surrounding basin was raised up to match sol II. In the second instance the structure was converted into a cistern with an aqueduct built leading to it. Thus, the open plaza of the agora seems to have continued, with the ‘granary’, even though the late portico had been destroyed: see appendix K7b for details. See also appendix K4b for a second late portico made of reused materials that might date from the reoccupation period and have continued intact during the 5th–6th c.
Dating summary (for construction of ‘granary’): range 378–403, midpoint 390.5, class Cs8 (contextual coins), publication 3/3.
Dating summary (for destruction of traditional architectural condition of agora): range 387.5–403, midpoint 395.25, class Cs8 (contextual coins), Cs9 (contextual pottery), Cs3 (associative, phase of development) publication 2/3 (as not all key coins and pottery published).
Dating summary (for reoccupation, at sol I levels): range 387.5–403, midpoint 395.25, class Cs8 (contextual coins), publication 3/3.
Dating summary (for reoccupation, at sol II levels, with first raising of square monument basin edge): range 387.5–403, midpoint 395.25, class Cs8 (contextual coins), Cs3 (associative, phase of development), publication 2/3.
Dating summary (for reoccupation at sol I level, with abolition of basin and aqueduct water supply): range 387.5–600, midpoint 493.75, class Cs8 (contextual coins), Cs9 (contextual ceramics), publication 2/3.
16AEG Cyrene: The evolution of the agora of Cyrene in Late Antiquity was complex. It began in the later 3rd to earlier 4th c. and extended over a number of phases. These phases often included the renewal of secular monumental buildings and the establishment of private houses at the same time. Thus, it is not easy to talk of abandonment, as the traditional functions seem to have carried on amidst unconventional building works. It is not always entirely clear how the established phases of the site were arrived at by excavators, although clearly there are a number of trends and changes in occupation levels, which do suggest that the phasing is broadly correct, even if the dating attached to these phases sometimes looks feeble. The risk is that the phasing has been simplified, in order to produce a coherent story, meaning that some sub-phases have been lost, especially because different parts of the agora have been published separately. It is difficult to offer a critique of the phases given in the report, as detailed drawings tend to focus on architectural remains, especially plans and architectural reconstructions, more than on the observation of stratigraphy and relationships. For this reason, we are obliged largely to accept the phasing offered in the reports, even if we can question some dating assumptions. One should note there is an important difference between the phases of construction, believed to be a response to periodic catastrophic events, and the periods of occupation, which extend between them. The reports are organised in terms of periods of occupation, although I am only interested in phases of construction.
In the first late antique construction phase, the first houses on the south side of the agora were built at a level thought to date to after the earthquake of AD 262: S. Stucchi and L. Bacchielli, L’Agorà di Cirene 2.4: il lato sud della platea inferiore e il lato nord della terraza superiore (Rome 1983) 111–13 [Guidoboni Catalogo no. 262]. Other houses of the same phase, in the south-west corner of the square, contained in their walls fragments of what the excavators identified as mid-3rd c. cooking ware. As there is more than one piece reported, we can take these ceramics to provide a weak contextual date of 250–75 for the construction phase of this encroachment: V. Purcaro, L’Agorà di Cirene 2.3: l’area meridionale del lato ovest dell’agorà (Rome 2001) 104. This is credible, as elsewhere in these reports such 3rd c. cooking ware is listed in terms of types recognised in the Athenian Agora: e.g. L. Bacchielli, L’Agorà di Cirene 2.1. L’area settentrionale del lato ovest della platea inferiore (Rome 1981) 185. However, the levels relating to the occupation of this first late phase also contained a hoard of coins of Constantius of AD 335–37 and Gallus Caesar of AD 352–54, a treasure that we should date as being deposited in the 25 years following 352: Stucchi and Bacchielli (1983) 111–13. Thus, we should envisage the occupation of this phase to extending a long time after its initial construction. Unfortunately, the stratigraphic record is not really published well enough to examine if building really did occur at a single moment, or if it extended over a longer period.
In the second late antique construction phase, the north-east corner of the agora saw further houses built, after some major disruption to the monuments of the city, in which the market building seems to have been ruined and in which spoliated temples were used as building material. It is important to note that these house encroachments occurred in the same phase as the last phases of public building that took place on the square: i.e. the construction of the last market building and the conversion of an adjacent temple into an audience hall (former Augusteion, called ‘praetorium’): see Stucchi (1965) 307–18, with appendix W1 for the market building. This period of this phase was thought by the excavator to date to after the ‘earthquake of 365’ (Amm. Marc. 22.16.4), partly based on the coin hoard evidence above, which suggested occupation of the previous phase up to at least 352: S. Stucchi, L’Agorà di Cirene 1: i lati nord ed est della platea inferiore (Rome 1965) 318–20. As there are no later finds reported in this phase, the coins of this hoard could generate a TPQ for the start of the next phase of 352, rather than rely on the earthquake. The hypothesis that the impetus for this phase of construction was an earthquake is not supported by any diagnostic observations of seismic damage or counter-measures (such as buttresses), even if there is evidence of rebuilding. Thus, I will not use 365 as a TPQ, preferring 352. For more commentary on the earthquake of 365, which I have not used here to provide a dating basis, see A. Di Vita, “Archaeologists and earthquakes: the case of 365 A.D.”, Annali di Geofisica 38.5–6 (1995) 971–76, as described by Amm. Marc. 26.10.16–19 and other authors, as listed in G. Kelly, “Ammianus and the Great Tsunami”, JRS 94 (2004) 141–67 [Guidoboni Catalogo no. 138], with further bibliography plus comments on Gortyn in appendix C4. For a TAQ for this phase we have only a theory, based on texts, that the next phase began ca. 400, as a result of imperial largesse, which is a poor date based on the assumptions involved.
In a third late antique construction phase, houses encroached further on the agora, reducing the plaza by up to half, now in a haphazard manner, so that the open space was no longer rectangular, and was effectively divided into three sub-courts. The ‘praetorium’ was also occupied by a house. However, the late market building was repaired in this phase: see appendix W1. This phase of construction, ‘ca. 400’, is believed to have been initiated as an imperial response to supposed Austurian destruction, following invasions attested at this time (recorded in numerous letters by Synesius). It is thought to be the result of successful intercession of the embassy of the prominent Cyrenean Synesius (Syn. Ep. 104, 113, 124) to the imperial court at this time (same letters in edn. of FitzGerald (1926) and Garzya (1989)). For most of the agora, there is nothing to confirm this dating, either in terms of construction or subsequent occupation. Unspecified dating evidence shows that the occupation of the houses on the agora continued until the mid-5th c.: Stucchi and Bacchielli (1983) 116–18, esp. 117 with fig. 80. Yet, there is more precise dating evidence for occupation on the northern side of the agora, where the floor of house 13, built inside the ‘praetorium’, produced a lamp of the beginning of the 5th c. (of Athenian Agora VII no. 2793 [which dates it as 5th c.] and Ephesos IV 2.103 nos. 968–995 [not seen], as described in Stucchi (1965) 330), a lamp that was also found on the floor in house 32, whilst the floor of house 33 produced a type of ARS, between A and B, then thought to belong to the beginning of the 5th c. (as described in Stucchi (1965) 336–37). These finds, from different places, suggest, as associative finds, an occupation period of the first half of the 5th c. This makes a construction date of ca. 400 for the beginning of the phase credible. However, this date depends on assumptions from texts, which do not actually record any building work as a consequence of the embassy: this is just an implication, given that it was successful. Thus, the date of building work in this third, late phase should be classified as poor. It is reasonable to locate the end of occupation, after this third phase of construction, around 450, based on associative finds, given the lack of material beyond this date.
Dating summary (first late phase of construction): range 250–75, midpoint 262.5, class Cs9 (contextual ceramics), publication 2/3.
Dating summary (second phase of construction): range 352–400, midpoint 376, class x (historical texts), class Cs8 (contextual coins), publication 3/3. Poor.
Dating summary (third late phase of construction): ca. 400, class x (historical texts), Cs3 (associative, finds), publication 3/3. Poor.
Dating summary (abandonment): ca. 450, class Cs3 (associative, finds), publication 3/3.
V5b Degradation and Disuse in the East, 5th C.
11THR Nicopolis ad Istrum: The agora was in decay in the 5th c., as revealed by pits and a drain through the agora. The pit fills containing as their latest coins examples of: 364–75 (latest of 9 coins, pit 1); 425–55 (latest of 9 coins, pit 2); and 395–408 (latest of 12 coins, pit 3), whilst the drain fill contains as its latest coin one of 455–56 (the latest of 115 coins, after examples of 408–50 and 425–35): P. Vladkova, “The Late Roman agora and the state of civic organization”, in The Transition to Late Antiquity on the Danube and Beyond, ed. A.G. Poulter (Proceedings of the British Academy 141) (Oxford 2007) (203–17) 209–10. See also P. Guest, “The Roman and Byzantine Coins excavated at Nicopolis ad Istrum and Gradishte, Bulgaria”, NC 159 (1999) (314–27) 314–19. We could insist on a contextual date for the beginning of decay based on the first pit, but the second pit with a larger coin sample provides perhaps the best basis for dating. Its finds suggest that the pit was dug and filled in sometime in the 25 years after 395. Cleaning finds coming from the surrounding streets, found between the paving stones were similar: of 21 coins, 1 was 3rd c. (Gordian II), 8 were 4th c., 6 were first half of 5th c., and the rest cannot be identified, except to fall in the 4th or first half of the 5th c.: Vladkova (2007) 208–209. However, the city endured, to be sacked by the Huns, probably in 447 (see appendix C10a). Thus, the decay of the agora probably represents a real change in the development of the city, rather than being representative of its development overall.
Dating summary (for beginning of decay of agora): range 395–420, midpoint 407.5, class Cs8 (contextual coins), Cs3 (associative, phase of development), publication 3/3.
10MAC Thessalonica: Clay pits for ceramic production have been found on the agora, with unspecified 5th c. pottery. Kilns have also been excavated, in and around the odeon-bouleuterion, which post-date its last major phase, including some that cut through the floor of the east portico: P. Georgakè and I. Zôgraphou, “The stratification of the square and the south-east sector”, in Αρχαία Αγορά Θεσσαλονίκης 1, ed. P. Adam Veleni (Thessalonica 2001) 65–86, English summary pp. 327–28; A. Valavanidou, “Workshops on the site of the ancient agora of Thessalonica”, in Αρχαία Αγορά Θεσσαλονίκης 1, ed. P. Adam Veleni (Thessalonica 2001) 121–30, with p. 125 fig. 6 and 126 fig. 7. English summary pp. 329–30.
Of dating, abandonment is placed at the very end of the 4th c. by Valavanidou (2001) and other reports in the same book, whilst industrial occupation occurs from the 5th c. (i) However, for the abandonment, we are given no contextual details to support a late 4th c. date. A TPQ for the abandonment can be derived from a coin found under a column base in a preceding phase. This coin provides a TPQ of 320 for both the rebuilding of the colonnade and for the mosaics, and thus ultimately for the abandonment that followed the rebuilding: V. Kalavria and A. Boli, “The stratigraphy of the east wing”, in the same volume pp. 39–64 esp. 47, with English summary pp. 326–327, esp. 327. A late antique bust, found in the south stoa of the agora for dating the forum occupation, could be used as evidence that the forum continued to be occupied at least until 380. This bust possibly dates from sometime in 380–460. The bust was used for a statue that already reused a Late Roman statue of possibly late 4th c. date (see appendix K9 plus LSA 2363). However, I have not been able to establish the bust’s display context properly, making it hard to be sure that it provides evidence of traditional forum occupation at this date. The only TAQ for abandonment seems to be derived from the subsequent industrial occupation, from the end of its dating range. (ii) For the industrial occupation, we have deposits relating to the conversion of the odeon into a productive area that have produced lamps of the late 4th to the first half of the 5th c. on p. 48, for lamps catalogued on p. 58 no. 30 and p. 63 no. 31. The first of these lamps is given as first half of the 5th c. and refers to J. Perlzweig, The Athenian Agora VII: Lamps of the Roman Period (Princeton 1961) 184 pl. 41 no. 2603. The second lamp is given as last quarter of 4th to first quarter of 5th c., referring to Perlzweig (1961) 55–56 without a Perlzweig lamp no., which seems to refer to a cross monogram of the end of the 4th to the first half of the 5th c. The lamp no. 2603 has the latest start date of the group, so gives a contextual date for the beginning of industrial occupation of 400–425. If these are use finds, the absence of materials dating after the mid-5th c. also suggests a TAQ of this date.
Overall, it has to be admitted that we do not have sufficient contextual detail to pinpoint the start of abandonment. However, it must be after 320, the coin-based TPQ of the colonnade rebuild. At the same time, we can date the industrial occupation, contextually, as being in the 25 years after the start date of its last identified ceramics, so to the period 400–25. This also gives us a TAQ of 425 for the abandonment. I am grateful to E. Rizos for help in checking my reading of the report.
Of subsequent phasing, it is interesting to note that shops adjacent to the agora, on its south side, carried on being occupied in the 5th–6th c., showing no sign of abandonment: see appendix Y6. It is troubling to see a statue base on the plaza that has a number of crosses on it—including one anchor cross and one cross with splayed ends on another side, large but not very well-drawn, alongside possibly a third cross on this side: L. Lavan site observation 2017. Crosses on public monuments tend to date no earlier than the first decade of the 5th c. They first appear ca. AD 400 at the beginning of civic inscriptions, in catacomb art, and on metalwork and ceramics: see dating foreword. It is possible, even likely, that the statue base does not originate in the agora and is just being stored here. But it is worth noting that the architectural survival of the agora is relatively good, given that it occurs inside a crowded city. Is it possible that the plaza had some period of re-occupation / part-reoccupation as a monumental square after a period of degradation, as seen in western sites such as Rome (Forum of Caesar), Ostia (Palestra), and perhaps Nora: see appendices K2a and V4b. Perhaps future excavations might be able to shed light on this.
Dating summary (for beginning of abandonment of agora): range 320–425, midpoint 372.5, class Cs7 (TPQ coin), Cs9 (contextual ceramics), publication 3/3.
Dating summary (for beginning of industrial use of agora): range 400–425, midpoint 412.5, class Cs9 (contextual ceramics), publication 3/3.
10MAC Athens (Athenian Agora): This square was greatly reduced in size, during the 25 year period after 395, with the construction of the Palace of the Giants and other structures: see I. Baldini Lippolis, “Sistema palaziale ed edifici amministrativi in età protobizantina: il settore settentrionale dell’Agorà di Atene”, Ocnus 11 (2003) 9–23) and appendix K2b. The new plaza was now rather rectangular than trapezoidal and a good deal smaller than its classical predecessor, having an open space of about half the size. The reduced area measured ca. 105 m east-west by ca. 109 m north-south. The agora was subsequently truncated in the third quarter of the 5th c., by the construction of an east-west colonnaded street, to a rectangle of ca. 57 m east-west by ca. 63 m north-south. Around this time a change of orientation is visible within the plaza, seen in the ‘Eudocia’ columnar monument. This structure is set against the alignment of the ‘rectangular’ agora side, so as to suit the direction of a new east-west colonnaded street, rather than that of the Panathenaic Way, which ran through it up to the Acropolis. This alteration in the major monumental route of the city must have reduced the status of the ‘early 5th c.’ agora. It was now a ‘side plaza’, which one drove past, rather than a plaza that had to be traversed by processions.
This change of status may have had an impact on its monumental development in the 5th c. The new east-west colonnaded street, of the third quarter of the 5th c., also reduced the size of the plaza, as the new colonnades extended a long way out into what had been the east side of the agora. However, all of these developments could still be seen as a modification of the monumental space, not a detraction from it. What followed after these developments can only be described as degradation, due to the appearance of industrial features within the plaza. The east side of the agora was now encroached by a leat [water channel leading to a mill] for a series of three water mills. The northernmost of these was set inside the agora ‘proper’, opposite the Square Building. This is best summarised by A. Frantz, The Athenian Agora, xxiv: Late Antiquity A.D. 267–700 (Princeton 1988) 80–81. Apart from changing the character of the plaza, the leat for the mills reduced the shrunken ‘early 5th c.’ agora yet again, by about one third of its size, to ca. 57 m east-west by ca. 63 m north-south, with a tapering exit towards the south-east for the Panathenaic Way. The effect was to make the surviving agora more trapezoidal than L-shaped.
Of dating, the ‘central mill’ of the three served by the water leat is dated to the third quarter of the 5th c., based on deposits beneath it. Here a build-up of water borne sand and gravel sealed coins (mainly late 4th to 5th c.), which stop in the reign of Leo (457–74), and “lamps and potsherds agree”. After this a watermill was constructed: see A.W. Parsons, “A Roman water-mill in the Athenian Agora”, Hesperia, 5 (1936) (70–90) 88. I take the ceramics plus the coin of Leo (which is the latest dated find, according to the limited information provided by the report) as indicating a contextual date of 25 years after 457 for the construction of the central mill. Of surrounding dating evidence, it is worth noting that the leat for the water mills covered over a wall that included in its foundations part of the Stoa Poikile, on the north side of the agora. This wall has been dated by late 4th and early 5th c. lamps (see appendix C5). On the east-west colonnaded street and its dating see appendix C3. Of the surrounding buildings, the adjacent Palace of Giants was in decay in the 6th c. or may have changed function, as revealed by a 6th c. rubbish deposit on the floors of basement rooms in the south-east of the complex (“the pottery indicates a date in the neighbourhood of AD 530”), whilst there is contemporary occupation evidence with later 6th c. lamps and Christian ampulae: H.A. Thompson, “Activity in the Athenian Agora 1960–1965”, Hesperia 35 (1966) (37–54) 45; Frantz (1988) 91. Likewise, isolated destructions and abandonments of 5th–6th c. date have been occasionally recorded in surrounding buildings (e.g. Frantz (1988) 82, Parsons (1936) 88, both 6th c.). However, neither sequence should distract us from the main issue: of what happened to the plaza itself, within a city that saw continued and complex occupation. Only the water mills are pertinent to this question.
Dating summary (for ‘monumental degradation’ by water mills): range 457–82, midpoint 469.5, class Cs9 (contextual ceramics), Cs8 (contextual coins), publication 2/3.
13ASI Hierapolis: According to brief and undetailed reports from Hierapolis, the Roman agora was destroyed in an earthquake in the second half of the 4th c. [no obvious candidates in Guidoboni Catalogo]. It was then left outside the fortification of the early 5th c. that was built on the south side of the site: M.P. Caggia, “Agorà Nord—Stoà Sud”, in Hierapolis di Frigia II. Atlante di Hierapolis di Frigia, edd. F. D’Andria, G. Scardozzi and A. Spanò (Istanbul 2008) 87. The area of the agora became an industrial area, notably with kilns for producing tiles in its northern area, around the west stoa: F. D’Andria, “Agorà Nord—Stoà Ouest”, (in the same volume) 86. There were also kilns for producing ceramics and rooms for different stages of ceramic production, in the northern area, around the east stoa: M.P. Caggia (2008) 87. P. Arthur, Byzantine and Turkish Hierapolis (Pamukkale) (Istanbul 2006) 109–118, describes the development in a bit more detail, without further chronology, noting kilns built after the late 4th c. earthquake surviving to the 6th or even early 7th c., producing household ceramics, including mortars, basins, jugs, amphoras, and tiles (p. 109). There are large quantities of waste dumps around the former agora surface (p. 110). Arthur also notes marble stacked up on the east side of the agora ready to be fed into lime-kilns (p. 111). He notes that most of the industrial activity in the area seems to have come to an end by the 7th c. earthquake on the site. I do not consider the occupation after the 7th c. here. Finally, a recent article on the fortification has claimed the earthquake that affected the agora area to be of the 3rd quarter of the 4th c., without substantiating this with evidence: L. Castrianni, I. Ditaranto and G. Di Giacomo, I. Ditaranto, G. Scardozzi (2010b), “La cinta muraria di Hierapolis di Frigia: il geodatabase dei materiali di reimpiego come strumento di ricerca e conoscenza del monumento e della città”, Archeologia e Calcolatori 21 (2010) 93–126. See also F. D’Andria, Guida Archeologica. Hierapolis di Phryrgia (Pammukale) (Istanbul 2003) 91–97 (not seen). I am left with little choice but to accept the dating range of all the articles suggested here, of 350–400, in the hope that in the near future the ceramic evidence behind the date will be published, as very likely the date depends upon this.
Dating summary: range 350–400, midpoint 375, class 0, publication 1/3. Poor.
See also (appendix V2): 13ASI Iasos in Caria: A cemetery was established over the agora, around a martyrion, whilst an encroaching room was built in the south stoa. These processes took place sometime in the 5th c.
See also (appendix V5a): 16AEG Cyrene: Encroachment by houses continued on the agora, reducing it to around half its size and effectively creating three sub-courts within the plaza, in the construction period dated ca. 400.
Discounted: 10MAC Messene: The agora was still occupied in the mid-4th c., when bases of honorific statues of this date are attested. After this time a thick layer of earth was deposited, before further activity of 6th–7th c. date: N. Tsivikis pers. comm. However, the deposition of this earth seems to be part of the wider transformation of the site into a village in the early 5th c. Therefore, the fate of the agora does not concern me here, as it does not inform us about the status of the agora in wider urban life.
V5c Degradation and Disuse in the East, 6th C.
10MAC Athens (Roman Agora): A cemetery with ‘Early Christian’ graves, found at the ancient ground level, was established at the Tower of the Winds, adjacent to the Roman Agora. It apparently did not extend into the agora itself: E. Tzavella, “Burial and urbanism in Athens (4th– 9th C. AD)”, JRA 21 (2008) 352–376, drawing on D.B. Small, “A proposal for the re-use of the Tower of the Winds”, AJA 84 (1980) 97–99, wishes to interpret this as evidence for ‘respect’ for the Roman Agora. Whilst this is a seductive hypothesis, the presence of a cemetery does not suggest a bustling commercial plaza. It could be seen as evidence of decay, if not obliteration, of the agora. Indeed, three ‘Early Byzantine’ funerary inscriptions seem to have been found around the agora, although their precise context is not certain, as far as I could see. These may indicate that a cemetery was started here in Late Antiquity: E. Sironen, The Late Roman and Early Byzantine Inscriptions of Athens and Attica (Helsinki 1997) 214–15, nos. 165–67, dated to the 5th–6th c. = IG (II/III, pt. 5 (2nd edn.)) nos. 13430, 13431, and 13438 (with plates XXVI and XXVII). The third inscription is now tentatively dated to the 6th c. by Sironen (based mostly on the general impression from the execution of the letters, e.g. delta and upsilon). Sironen uses a generic 5th–6th c. date for Christian funerary inscriptions in Greece, following D. Feissel, “Inscriptions du IVe au VIe siècle”, in D. Feissel and A. Philippidis Braat, “Inventaires du Péloponnèse”, TravMém 9 (1985) (267–395) 358–59. Feissel notes the absence of clear 4th c. Christian funerary inscriptions and the advent of radical change in the nature of urban communities after the end of the 6th c. as justifying his 5th–6th c. dating for these inscriptions. This principle is followed by Sironen, in cases where there is no evidence to the contrary: E. Sironen pers. comm., November 2016. There are also crosses at the start of Sironen (1997) nos. 165 and 167. The use of crosses is unlikely before ca. 400, based on the appearance of the symbol at the start of honorific inscriptions, in catacombs, on ceramics and on metalwork: see comments in foreword on dating. All that can really be said is that there is no stratigraphic evidence available for the Roman Agora. I have not seen M. Hoff, The Roman Agora at Athens (Ph.D. diss., Univ. of Boston 1988).
Dating summary: 400–600, midpoint 500, class z (regional development), Cs1 (artistic style), publication 3/3.
10MAC Corinth: Evidence for the degradation and disuse of the agora is not straightforward, especially as most of the agora was not documented in a stratigraphic manner. The fate of the agora is important however, as the churches of Corinth indicate that the city continued to be a place of some standing during most of the period.
Of positive building work, the latest seems to be the creation of the great staircase across the southern side of the agora, along with two fountains, which can be dated to 350–75, based on contextual coins (appendix K6), or the restoration of two porticoes suggested by architrave inscriptions of Valens and Valentinian I of 364–75 (appendix K2b). There are other slight indications of continued occupation in the area. A cross scratched on one of the pillars of tribunal suggests that it persisted into the 5th c. (see appendix K6). A date of after 400 would be appropriate for this engraving, based on observation of the display of crosses in the catacombs and at the start of honorific inscriptions (see appendices C3 and F3). The finding of a group of late chlamys statues from the agora seems to indicate late occupation, but attempts to date them within Late Antiquity look ill-founded: see appendix K9.
Of degradation, we have three areas of evidence:
(i) On the west side of the agora an encroaching rectangular building was constructed, either directly onto the paving or over the cement bedding of the paving [presumably where slabs had been robbed]. This building is aligned with a water channel that was likely connected to it, as the orientation of both structures was odd in relation to the surrounding buildings. This channel, which used spolia from the façade of the west shops in its construction (fragments of column drums) contained, in its fill, pottery of the mid-4th c. and also 14 coins from 367 to 425: C.K. Williams II, J. MacIntosh and J.E. Fisher, “Excavation at Corinth, 1973”, Hesperia 43.1 (1974) (1–76) 7–9. These finds suggest that the drain fell out of use no later than the period 425–50, with the building being established sometime before that, probably in the 25 year period following the first coin lost in the drain, so 367–92. The excavators believed that the channels’ use of building material from the west shops implies that it was constructed after an earthquake of 365 or 375, but it could be earlier [Guidoboni Catalogo nos. 138 and 143].
(ii) This same building implies the partial robbing of the paving in this area, whilst its replacement in road metalling is mentioned at least for the area east of the new building.
(iii) We also have evidence of spoliation and construction on the agora surface. Spoliated blocks taken from the buildings of the west terrace, were found “in the network of early Christian walls which began to spread over the edges of the Agora area” [such as the shop façade elements mentioned above], according to R.L. Scranton, Corinth, Vol. 1 Part 3: Monuments in the Lower Agora and North of the Archaic Temple (Princeton 1951) 73, who notes that 6th c. coins were found “in conjunction with some of them”, which provides a rough associative date of this century for this process of overbuilding, which looks like the functional abandonment of at least part of the agora.
Of decay / abandonment of the traditional function of the agora, we have three indications:
(i) The presence of a deposit, excavated on the pavement next to the podium / bema on the south side of the agora, extending in all directions around it, suggests that the paving of the forum was no longer being cleaned. One coin of Arcadius (395–408) and two coins of Anastasius (491–518) were found in this deposit, which was a homogeneous layer of red earth. This layer contained fallen roof tiles and “sherds exclusively of Late Roman date”: E. Ivison, “Burial and urbanism at late antique and Early Byzantine Corinth (c. AD 400–700)”, in Towns in Transition: from Late Antiquity to the Middle Ages, edd. N. Christie and S.T. Loseby (Aldershot 1996) (99–125) 104, quoting unpublished Corinth notebook 158 (1936) 166, 172 and Corinth notebook 159 (1936) 225–26, 243, 301. Ivison takes the coins as indicating the chronological range of the layer. M. Rothaus, Corinth: the First City of Greece, an Urban History of Late Antique Cult and Religion (Leiden 2000) 26 n. 33, identifies the coin of Anastasius as a TPQ. He also suggests that this is not a good TPQ, as coins of Anastasius are the most commonly lost at Corinth, a remark which does not make stratigraphic sense. Under the rules of this study, the deposit should be evaluated differently. The homogeneous nature of the deposit does not suggest that it was formed over a long time period, so a contextual date is appropriate. A single coin of Arcadius, which could still be in circulation at the time of Anastasius, does not provide a basis for dating, but rather the coins of Anastasius, as the latest published finds from the deposit should provide a basis for its deposition date, of 25 years after their start date, so 491–516. Given the lack of detail on the ceramics, it seems best to disregard their attribution as ‘Late Roman’.
(ii) The spread of the possibly 6th c. buildings, mentioned above.
(iii) The cemetery that develops in this area, described by Ivison, does not seem to encroach over most of the plaza itself, only around its edges. The cemetery did eventually ring the plaza on three sides, as shown on Ivison’s plan on fig. 5.1 on p. 100. Dating the beginning of the burials is, however, difficult. It is clear from Ivison’s review (pp. 110–111) that the possibly 5th c. funerary structures, over the South Basilica area, are only dated by stylistic analogy with mosaic burials in other Greek cities, which I do not wish to explore here. However, a few burials actually within the plaza, on the south-west corner of the forum, seem to be better dated, as Ivison (1996) 110 notes, so I will prefer these. A double-vaulted chamber excavated in 1974 contained a ‘Lekythos’ dated to the late 6th c.: C.K. Williams and J.E. Fisher, “Corinth 1974: Forum Southwest,” Hesperia 44 (1975) (1–50) 16–17, with dating depending on Robinson Athenian Agora V, M-367 p. 118 pl. 34 [from the late 6th c. layer in Robinson]. A vaulted tomb with enclosure wall, excavated in 1973, produced a wheel-ridged oinochoe, which is compared with H.S. Robinson, Athenian Agora V: Pottery of the Roman Period, Chronology (Princeton 1959) M-365 p. 118 with pl. 33 [late 6th c. layer in Robinson] and N-8 p. 122 with pl. 35 [early 7th c. in Robinson], given by Williams and Fisher as being of probably 7th c. date. Furthermore, Williams and Fisher (1974) 10 state that “conclusive stratigraphic evidence suggests a date for the construction of the tomb of no earlier than the second quarter of the sixth century after Christ and slightly later”. The foundation trenches of the tomb walls, alongside the floor of the tomb enclosure, produced coins that rule out a date before the 5th c.: Williams and Fisher (1974) 10. Quite what the “conclusive stratigraphic evidence” could be is not clear from the report, although the vault does overlie the ruins of the 4th c. encroachment building south of Temple F, described above. Ivison (1996) 110, drawing on this evidence, maintained that stratigraphic evidence and jugs associated with the forum burials date the appearance of multiple burial vaults no earlier than the second quarter of the 6th c. However, this looks to be a simplification. The clearest (directly associative) dating evidence comes from the tomb excavated in 1974, where a ‘lekythos’ of likely late 6th c. date was identified from the contents. Otherwise, we only know that the 1973 tomb post-dates 400. I am not given access to the subtle stratigraphic arguments in favour of a date in the second quarter of the 6th c.
Overall, it looks like there is some reason to believe that the occupation of the agora as an open space continued into the 5th c., without significant decay before this time. Some degradation, like encroachment and the removal of paving stones, appeared already in the 4th c., first dated in the period 367–92. Full blown decay, manifested by the clay surface around the bema or large-scale overbuilding, is dated to 491–516 and to the 6th c. respectively. Further nuances may have been lost in the rough excavation of the agora. Yet, at present, the datable evidence suggests that the loss of the agora as an open space is best situated in the 6th c., when overbuilding filled the area and a cemetery surrounded it. Whilst the cemetery certainly began at some point after AD 400, it seems perverse to ignore that all the evidence from within the properly-excavated tombs on the agora actually dates to the late 6th c. and early 7th c., so the period 587.5–612.5. Thus, it seems fair to give the beginning of the cemetery a date within this 25 year period on grounds of associative finds.
Dating summary (beginning of degradation, via encroachment of paving): range 367–92, midpoint 379.5, class Cs8 (contextual coins), publication 2/3.
Dating summary (beginning of decay): range 491–516, midpoint 503.5, class Cs8 (contextual coins), publication 3/3.
Dating summary (beginning of abandonment, with overbuilding): range 500–600, midpoint 550, class Cs3 (associative, coins finds), publication 2/3.
Dating summary (beginning of cemetery): range 587.5–612.5, midpoint 600, class Cs3 (associative, ceramic finds), publication 2/3.
13ASI Xanthos (in Lycia) (Theatre Square / Roman Agora): Abandonment took place sometime before the late 6th c. to mid-7th c., when a cemetery was established across the square with finds of this date (unspecified ceramics and dress ornaments—earrings, pendentive crosses, bracelets and rings—with an apparent parallel to a hoard from Palmyra of the very end of the 6th to 7th c.). A church was set into this cemetery, during its phase of use, thus sharing the same dating as it (see appendix V2). This church was constructed in one corner of the square and includes spolia thought to come from the plaza: J. des Courtils and D. Laroche, “Xanthos—Le Letoon. Rapport sur la campagne de 1998”, Anatolia Antiqua 6 (1998) (457–77) 463–68. A tomb (no. 22), covering the demolished walls of the church, uses a piece of the entablature from the surrounding square. It has been suggested that other tombs in the square, underneath the church, date to the 5th c., but these have not so far produced datable finds, as far as the reports make clear: J. des Courtils and D. Laroche et al., “Xanthos—Le Letoon. Rapport sur la campagne de 1998”, Anatolia Antiqua 7 (1999) (367–99) 373–76.; Ibid. (2000) “Xanthos—Le Letoon. Rapport sur la campagne de 1999”, Anatolia Antiqua 8 (2000) (399–83) 344. It is possible that the surrounding square remained partially intact whilst the church was present. I saw repairs in reused material to a pier in the colonnade of the square (replaced in other parts by a rubble wall), which suggested that it could well have been repaired after an act of damage which produced some spolia: L. Lavan site observation 2005. Thus, we should disregard the significance of the architrave used in tomb 22. Its reuse might not have spelled the end of the portico. The change we can be sure of is one of function, from a plaza to a cemetery, implying the abandonment of the traditional function of the plaza. We cannot be sure of the chronology of architectural degradation.
Dating summary (for abandonment): range 587.5–650, midpoint 618.75, class Cs9 (contextual ceramics), Cs8x (contextual, other objects), publication 2/3.
13ASI Ephesus (Upper Agora): The decay of the Upper Agora is a complex subject. The picture given by W. Alzinger, “Das Regierungsviertel”, ÖJh 50 (1972–75) (Beiblatt 229–300) 295; W. Alzinger, “Ephesus”, in RE Supplement 12, 1634–36 and C. Foss, Ephesus after Antiquity: a Late Antique, Byzantine and Turkish City (Cambridge 1979) 188–91 are now out of date. For much of Late Antiquity, the agora seems to have continued: the east portico was rebuilt (appendix K4b) and the civil basilica was restored, sometime before ca. 491–516 (see appendix X1b). A bath building on the north-east side of the agora also saw late repairs, with a 5th c. mosaic inscription (in a mosaic over 40 m long) recalling the efforts of the proconsul Asklepi(o)s: IvE 4.1313; H. Vetters, “Ephesos. Vorläufiger Grabungsbericht 1971”, AnzWien 109 (1972) (1–120) 85 (plus plates 2 and 3). It is important to note that the temple in the centre of the agora, which was demolished and levelled at some time, cannot now be linked to any other late antique developments in the plaza (such as the supposed higher occupation level in the east portico). Neither can its demolition be dated to Late Antiquity. Probably it is best to separate degradation from abandonment, as changes in monumental buildings might have taken place whilst the plaza was still open and used. It is notable that the odeon-bouleuterion does not seem to have experienced demolition or spoliation, in comparison to nearby structures.
Of degradation within Late Antiquity, we can note three changes: (i) At some point in Late Antiquity honorific monuments have likely been removed from the plaza, which is strangely empty of statue bases, compared to the adjacent Embolos. This is implied by the silence of M. Aurenhammer and A. Sokolicek, “The remains of the centuries. Sculptures and statue bases in late antique Ephesus: the evidence of the Upper Agora”, in Archaeology and the Cities of Asia Minor in Late Antiquity, edd. O. Dally and C. Ratté (Kelsey Museum Publication 6) (Michigan 2011) 43–66. This absence of late antique statue bases from the Upper Agora, in contrast to other areas of the city, is noticed by S. Ladstätter and A. Pülz, “Ephesus in the Late Roman and Early Byzantine period: changes in its urban character from the third to the seventh century A.D.”, in The Transition to Late Antiquity on the Danube and Beyond, ed. A.G. Poulter (Proceedings of the British Academy 141) (London 2007) (391–433) 404. It has to be admitted that ‘editing’ of the statue landscape in a plaza is attested elsewhere, both in Early Imperial and late antique context (as at Ostia, Constantinople and Sagalassos, see main text), and did not in these cases represent a terminal degradation of the plaza, only a reform. However, in the case of the Upper Agora of Ephesus, the clearance seems to have been wholesale, leaving nothing behind, which does suggest degradation. (ii) The basilica that defined the north side of the square was degraded and then demolished, after having being rebuilt within Late Antiquity. It also underwent a Christianisation, involving the erasure of the name of Artemis from the dedicatory inscription and the carving of crosses on the heads of portrait busts of Augustus and Livia, which were later buried here. This occurred sometime after 400, but before 516, based on artistic parallels for the use of crosses on public monuments / private material culture, and on a contextual coin date from a tile floor that sealed the burial of the busts. This tile floor represented a degradation of the basilica, perhaps its conversion into shops, as it only covered a small part of the structure: see appendix X1b. A piece of the basilica’s dedicatory inscription [alongside capitals and bases] was found at the church of St. John. This provides a tentative TAQ for the basilica’s destruction of the death of Theodora in 548, whose monograms adorn this church, alongside those of Justinian: W. Alzinger (1972–75) 266 and 299. We do not yet know the archaeological context of these fragments, but the translocation of these pieces to this church, which is outside of the city, suggests an association with a major campaign of building work, probably the early 6th c. church: see appendices X1b and also K2b (Tetragonal / Lower Agora) for a list of other pieces incorporated in mid-6th c. building work. (iii) A late possible ‘house’ was built in the north-east corner of the square, incorporating columns from the upper storey of the basilica. It is not actually set in the main hall of the structure but in an adjacent room, at its eastern end. The pieces used in this building work perhaps arrived there during the late rebuilding works to the civil basilica, and so do not relate to the decay of the structure (see appendix X1b).
Of abandonment, a more obvious indication of a change of function comes in the form of ‘Byzantine’ domestic structures built over the plaza itself, with a related water supply, a development that is paralleled in the Prytaneion, where similar structures are built over it, though this does not seem itself to be dated: F. Eichler, “Die österreichischen Ausgrabungen in Ephesos im Jahre 1964”, AnzWien 102 (93–109) 96f. See appendix Q1 for the poorly- dated structures inside the prytaneion and K2b on the reuse of architectural parts from this building and others in the second half of the 6th c.
Overall, it seems that the degradation of the basilica cannot yet be given a date before 491. It was obviously substantially intact in the years following 400, in order to be Christianised, notably with inscribed crosses. There may have been some degradation before 491, as the building is vast, but we have no evidence for this. The process had certainly begun by 516. We cannot know if the degradation of the basilica equated with the degradation of the rest of the agora plaza, although it serves as a pointer, in the presence of other undated changes. Given that the basilica made up the whole of the north side of the agora, serving as its portico, the impact of changes to it on the appearance of the agora will have been considerable. The abandonment of the agora plaza itself is not well dated, but we have no reason to place it before 491. The process of abandonment, with over-building by domestic structures, does seem likely to have occurred before a well-attested reduction in intensity of occupation, in the area south of the late city wall, seen during the reign of Heraclius (610–41): coins recovered from this post-antique ‘extramural zone’ do not include the relatively common issues of Constans II (641–68): see appendix H7 (preliminary comments on the statue monuments of the Embolos). This gives us a TAQ of 641 based on site development.
Dating summary (degradation of agora): range 491–516, midpoint 503.5, class Cs8 (contextual coins), publication 3/3.
Dating summary (abandonment of agora): range 491–641, midpoint 566, class Cs8 (contextual coins), z (site development), publication 3/3.
See also (appendix A11): 15ORI Gerasa (Oval Plaza): Encroachment of the northern tip of the plaza, sometime between 387.5 and 602, as part of replanning of the shops of the adjacent north-south street.
See also (appendix K2b): 13ASI Laodicea ad Lycum (North Agora): After destruction in the earthquake of AD 494, the site saw occupation by brick-floored structures, including a spolia workshop and a small church, in the period 494–610.
See also (appendix S2) 10MAC Dyrrachium: The complex is dismantled from 570–600 and then graves of unverified 7th–8th c. date cover the plaza.
V5d Degradation and Disuse in the East, 7th–8th C.
13ASI Aphrodisias: Coins in the North Agora extend throughout the 6th c., into the 7th c., with one coin of Phocas (reigned 602–10) and three of Heraclius (reigned 610–41). In contrast, there is one coin of the relatively common issues of Constans II (reigned 641–68). There is then a clear break in the intensity of occupation, with wind and water deposits, before Mid-Byzantine occupation: C. Ratté, “The urban development of Aphrodisias in Late Antiquity”, in Urbanism in Western Asia Minor: New Studies on Aphrodisias, Ephesos, Hierapolis, Perge and Xanthos, ed. D. Parrish (Portsmouth, Rhode Island 2001) (117–47) 138–39, 145. The archaeological context of the coins on this open space, which was never paved, is not made clear, and is perhaps not known, as early excavations in the city under Kenan Erim were rough. In the absence of further details, we will take the coins as representing use-finds lost in the agora, rather than as finds relating to the fill of pits or from hoards. Under the rules of this study, I would place the abandonment of the agora in the 25 years after the start date of the last coin find, which seems reasonable given the low number of coins of Constans II, which are relatively common in Asia Minor.
Dating summary (abandonment of agora): range 641–66, midpoint 653.5, class Cs8 (contextual coins), publication 2/3 (as context not clear).
13ASI Sagalassos (Upper Agora): The latest floors in the macellum include unspecified finds that appear to date from 487.5–575, based partly on subsequent stratigraphy, whilst the latest occupation layers in the southern part of the western portico shops (appendix Y2) contained two early 7th c. coins (after 602 and perhaps 631–40), suggesting that occupation on the agora continued until 602 or 631.
Decay is indicated by three final deposits in the agora: (i) the dumping of refuse inside the subdivided rooms of the ‘North-East Building’ in the same last pottery phase (phase 9): Internal Report (2004) 8–14; (ii) the dumping of refuse inside the macellum (dated on phase 9 ceramics, AD 550/75–ca. 700 in 2013); (iii) the filling of a late basin in the north-west corner of the plaza with domestic refuse (including ceramics of the late 6th to 7th c., probably phase 9 ceramics): Internal Report (2000) 42. Whilst the first two dumps only really indicate the disuse of individual buildings, the third is more indicative of the condition of the plaza, as it comes from a low basin, right by the nymphaeum, that probably was well-suited to casual water-getting. Without specific dated wares we must use the full range of phase 9 ceramics as our contextual date for the start of decay.
Destruction has been proposed in part of the agora. In the northern half of the western portico destruction debris was found, which was overlain by a spolia wall that reinforced the apse of the Bouleuterion Church, without taking the time to clear out the destruction debris first: Internal Report (2000) 27 and 33–35 (layer 5 in the east part of Room B). This repair perhaps followed an earthquake, although I would not like to speculate on the date of this. Strangely, no finds are reported from the destruction layer, which is a loose layer full of tiles, stones and mortar, found in other rooms elsewhere in the same part of the portico (p. 41, layers 2 and 3 in adjacent portico). A more recent study seems to confirm a lack of finds in the destruction layers, but suggests rather they are the result of gradual collapse rather than an earthquake, noting that this north part of the portico shop was abandoned by the end of the 6th c., based on ceramics and coins, in contrast to the southern half of the row: T. Putzeys, Contextual Analysis at Sagalassos: Developing a Methodology for Classical Archaeology (Ph.D. diss., KULeuven 2007) 330. Rather than date the destruction to 650, the date of a known earthquake, as do the excavators [not in Guidoboni Catalogo], I prefer just to place this event after the start date of the last occupation finds in the rooms of the northern half of the west portico: phase 8 ceramics (AD 450/75–550/75) and 4th–5th c. coins, plus one of Justinian (reigned AD 527–65): Putzeys (2007) 294 with Internal Report (2000) 19 (the reference to the Justinianic coin is in Putzeys alone). Phase 9 ceramics were notably absent. Thus, the abandonment is likely to have happened within the period defined by the start of the coin of Justinian of 527 and the upper limit of phase 8 of 575, with the destruction probably no later than 25 years or so after that, or one might expect some phase 9 to have entered the layers, of which there is none. There seems no basis for a mid-7th c. destruction of the bouleuterion church, from chasing up references, only belief in a mid-7th c. earthquake.
Abandonment can be suggested from the absence from the agora of coins of Constans II (reigned 641–68), which are relatively common in Asia Minor. This implies abandonment of the agora in the range of 25 years following 631, the start date of the last find (though with uncertainty over 602 or 631 as a start date). In contrast, coins of Constans II are present on the Lower Agora of the same city.
Dating summary (for start of decay): range 550–700, midpoint 625, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (for ‘destruction’ of west portico shops): range 527–75, midpoint 551, class Cs8 (contextual coins), publication 3/3.
Dating summary (for start of abandonment): range 631–56, midpoint 643.5, class Cs8 (contextual coins), publication 3/3.
13ASI Sagalassos (Lower Agora): Destruction deposits have been found on both sides of the agora marking the end of the plaza. Occupation deposits from the cellular units / shops established inside the west portico contained coins up to the reign of Phocas (AD 602–610), as well as pottery of Sagalassos Red Slip Ware phase 9 (which was AD 550/75–ca. 700 in 2013): see appendix Y2.
East side: Destruction deposits were found on the agora monuments that were cut by a water pipe, which passed over the ruins of the east portico and its rooms: see Internal Report (2003) 95–96 with p. 93 fig. 34. The late ‘guard rooms’ at the north end of the east portico rooms were also filled by a destruction layer (layer 2) containing “7th century AD material”: see Internal Report (2003) 97. However, Putzeys has recently clarified that the destruction levels in both ‘guard rooms’ contained phase 9 ceramics and that the adjacent structure (‘thermopolion’) contained phase 9 ceramics and some mid-7th c. coins (the latest being Constans II (AD 641–68): see appendices Y3 and S4.
West side: At the Agora Gate, a destruction layer (layer 4, in other areas layer 2) lies on top of the 6th c. AD stairway (severely damaged by falling debris from the collapsed western podium, so suggesting an earthquake [not in Guidoboni Catalogo]), which contained pottery dated to the first half of the 6th c. AD or later. Layer 4 contained several coins, dated between the middle of the 4th and the beginning of the 7th c. AD (the latest being AD 603–604). Over this earthquake debris a cemetery developed, covering the whole area, across the 6th c. stairway and beyond. Three tombs in this area contained coins (of the 4th– 5th c. and of the reign of Heraclius (613–15)): M. Waelkens et al., “The 1994 and 1995 excavation seasons at Sagalassos”, in Sagalassos IV. Report on the Survey and Excavation Campaigns of 1994 and 1995, edd. M. Waelkens and J. Poblome (Acta Archaeologica Lovaniensia Monographiae 9) (Leuven 1997) (103–216) 208–10. However, more recently extensive excavation of this cemetery has revealed that it dates to the 11th–13th c., based on radiocarbon dating and analysis of the finds: S. Cleymans and P. Talloen, “Protection in life and death: pendant crosses from the cemetery of Apollo Klarios at Sagalassos”, in European Journal of Archaeology (2017) (forthcoming).
Overall, the collapse layer at the Agora Gate (which was not cleared away) is the most credible piece of evidence for the destruction. However, the violence of this event, with falling stone pieces, clearly makes it probable that the destruction layers on the east side of the agora also date from the same time. Thus, I take the finds from these, which are slightly later than the west side / Agora Gate finds, as indicating a date in the 25 year period after the start date of the latest finds (so 641–66, based on the coins from the shops, with the coins allowed to take precedence over the ceramics, given their tighter dating, which does not contradict the ceramics).
Dating summary (for the destruction): range 641–66, midpoint 653.5, class Cs8 (contextual coins), Cs9 (contextual ceramics), publication 3/3.
15ORI Hippos / Sussita: A number of changes took place in Late Antiquity, which can be classified as functional rearrangements of the forum, although other changes look more like degradation: A. Segal et al., Hippos-Sussita. Fourth Season of Excavations: June–July 2003 (Haifa 2003) 5–6; A. Segal et al., Fifth Season of Excavations (September–October 2004) and Summary of All Five Seasons (2000–2004) (Haifa 2004) 13, 17–18, 29–31, 52–72 with figs. 7–8 for plans, which are essential for understanding the text.
Of traditional classical monumental investment in the plaza, we have a late staircase, which connected the forum to an adjacent square (‘the Hellenistic compound’) to the north: Segal et al. (2003) 5. See also Segal et al. (2004) 31 with figs. 7–8 for plans. The compound had perhaps held a series of temples, one Hellenistic, then one Roman, as suggested by architectural fragments: Segal et al. (2004) 29–30. Significantly, the portico columns of the plaza seem to have survived unspoliated until the end of the site in AD 749: Segal et al. (2003) 9. This implies that the architecture of the main plaza was still intact during the Umayyad period, which is itself very significant.
Of functional rearrangements, shops and houses were built on the north and east sides of the forum: Segal et al. (2004) 17. Late cellular shops occupied the ‘exedra’ / portico that marked the boundary between the two plazas, whilst a church was established on the northern edge of the ‘Hellenistic compound’: Segal et al. (2003) 24–33; (2004) 83–96. Wine or oil presses associated with the new church: Segal et al. (2004) 22–23 and 143–44. The ‘shops’ established on the north side of the forum, inside the ‘exedra’ / portico, seem to be ca. 2.5 m deep internally, whilst the only room division recovered defines a space that is internally 3.6 m wide. The row is at least ca. 13 m long, hand-measuring off the plan. I could not find any information about the houses installed on the east side of the plaza.
Of degradation, very carelessly built ‘Umayyad’ walls were exposed on the forum in the north-west corner, along with two millstones, in a context dating from this period. This implies some degradation of the plaza’s monumental appearance and function: A. Segal et al., Sixth Season of Excavations (July 2005) (Haifa 2005) 8.
Of destruction, we can note a 6th c. event that destroyed the shops, probably in 570–95 (see below). However, the end of the site as a whole is attributed to the earthquake destruction of AD 749 [Guidoboni Catalogo no. 220], which caused catastrophic damage across the site (Segal et al. (2003) 9) and also in the adjacent cities of Pella and Scythopolis, where it is well-attested archaeologically: see appendices A8b and A11. In terms of diagnostic seismic evidence for this destruction, we can note that the portico colonnades fell flat over onto the plaza (Segal et al. (2003) 9) and that there is evidence across the site for collapsed buildings with upper storey contents preserved that are not covered by any earth (Segal et al. (2004) 135–36). It is likely that there is also a general end of occupation at this time, which the excavators feel ought to correspond with this local but very powerful event.
Of dating of these phases, we have several indications:
(i) The staircase (of which the foundation of reused blocks survive, with two steps in one place): “The pottery, building materials, and construction methods all testify that both the pavement and the staircase were built during the Byzantine period”: Segal et al. (2003) 5. ‘Byzantine period’ in Israel usually means ca. 400–635, with ‘Late Roman’ being the period 200–400.
(ii) The construction of the shops on the north side of the forum, in the ‘exedra’, is thought by the excavator to date to sometime shortly after AD 400, on the basis of pottery found beneath the floor (4th c., with the exception of the latest sherd, a cooking ceramic of Adan-Bayewitz Competing Form C4B [not checked by me] (given as latter 4th to latter 6th c., which I take as implying late 4th to late 6th c., so AD 387.5–600, rather than my own use of ‘latter’ to mean second half of a century): Segal et al. (2003) 59 (wares 13–16, which are given on p. 58 as coming from under the floor). Under the rules of this study, this gives me a contextual construction date for the shops of 387.5–412.5.
(iii) The occupation of the shops can be traced from pottery above the shop floor, which ends in the 7th c. AD with LRD/CRS 10A [ca. 625–675 in LRP, but now considered the same as Hayes 9C in Atlante p. 239, which is ca. 580/600–700 in LRP and Atlante], LRC form 3 [ca. 570–660+ in LRP] and LRC form 10a [570–612.5 in Atlante]: J. Mlynarczyk, “Pottery report”, in Hippos-Sussita. Fourth Season of Excavations June–July 2003, A. Segal et al. (Haifa 2003) (50–75) 51–52, 58–59. A later sample of pottery from within the shops, under their collapsed roof, collected in 2004, includes 4th–5th c. pottery but also a casserole of the ‘Byzantine / Umayyad’ period and finewares of the mid-5th to early 7th c. date (imitation of ARS Hayes 91A–B given as ca. mid- 5th to early 6th c. [Hayes 91 is 350–650 based on LRP, not contradicted by Atlante or Bon, who lists the many variants with shorter date ranges], LRC Hayes 10a, given as late 6th–early 7th c. [570–612.5 in Atlante], CRS form 2 early versions, given as ca. AD 450 [400–550 in LRP], and same form later version, given as early mid-6th c.): J. Mlynarczyk, “Pottery Report (Hippos-Sussita 2004)”, in Fifth Season of Excavations (September–October 2004) and Summary of All Five Seasons (2000–2004), A. Segal et al. (Haifa 2004) (140–163) 142 and 150. From these wares, the LRD/CRS 10a seems to have the latest start date, suggesting that the occupation ceased in the 25 years after 580, so 580–605. Some sort of violent disjuncture is likely, perhaps an earthquake, as the roof collapsed. However, this was not the earthquake of 749 [Guidoboni Catalogo no. 138], which seems elsewhere to mark the end of the site, as here we have later occupation.
(iv) The church was not initially dated, although the 5th–6th c. has been suggested by the excavators: Segal et al. (2004) 52–72 esp. 68. However, a recent sondage under the atrium revealed a fill containing, as its the latest sherds, Hayes LRC form 3D, given as late 5th c. [487.5–500 in LRP], or form 3E, given as the “5th/6th century” [487.5–500 in Atlante]: J. Mlynarczyk, “Pottery report”, in Sixth Season of Excavations (July 2005); A. Segal et al. (Haifa 2005) (113–39) 119. I would thus take the construction period of the church as being in the 25 years after the start date of the latest ceramic, so AD 487.5–512.5.
(v) The winery adjacent to the church has produced an assemblage, functionally and chronologically associated with it, which has, as its latest datable fineware ceramics, CRS/LRD form 9B, given as of 580/600 to 700 [ca. 580–700 Atlante], with other fineware and other pottery beginning ca. 550 (CRS/LRD form 11, given as 550–650 [ca. 550–650+ in LRP] and ARS 104C given as ca. 550–625 [550–650 in Bon]) and some local and cooking wares, which are first known elsewhere in strata of AD 650–750: Mlynarczyk (2004) 143, 153–54. As this is a ‘use deposit’ for the winery, the occupation is likely to have started no later than ca. 625, as this is the lowest end date of the sherds listed, and no earlier than 550, based on the earliest start date. However, the coarsewares suggest a slightly later date, pushing me to suggest that the occupation began in the period 625–650. The ceramics are ‘functionally associated’ with the activity, meaning that they were used in and around the pressing process, rather than being from a dumped fill.
(vi) The very carelessly built ‘Umayyad’ walls in the north-west corner of the forum were associated with a deposit, containing two millstones, which seems to have produced “Byzantine and Umayyad” pottery (context F-370, no further details given): Segal et al. (2003) 6. The presence of ‘Byzantine’ pottery obviously problematizes a start date for the walls in the Umayyad period, as implied above. I would prefer to attribute these walls to the period after the destruction of the shops, which seems to have occurred in 580–605, on grounds of site phasing, although here I follow logic more than available evidence. The severe earthquake of 749 is used here as a TAQ, on which see appendices A11 and S3.
Dating summary (establishment of staircase): range 400–635, midpoint 517.5, class Cs9 (contextual ceramics), publication 2/3. Generic late antique date.
Dating summary (establishment of shops): range 387.5–412.5, midpoint 400, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (establishment of church): range 487.5–512.5, midpoint 500, class Cs9 (contextual ceramics), publication 3/3. Exemplary.
Dating summary (establishment of winery): range 625–650, midpoint 637.5, class Cs9 (contextual ceramics), publication 3/3. Unconventional.
Dating summary (destruction of shops): range 580–605, midpoint 592.5, class Cs9 (contextual ceramics), publication 3/3.
Dating summary (establishment of milling area and rough walls, marking degradation of site): range 580–749, midpoint 664.5, class Cs9 (contextual ceramics), z (site phasing), Cs2 (catch-all other, earthquake), publication 2/3.
Dating summary (destruction of site): 749, class Cs2 (catch-all other, earthquake), publication 1/3.
15ORI Gerasa: The round plaza seems to have still existed at the time of the Umayyad mosque, with the tetrakionion intact. However, a platform of the first phase of the mosque (of ca. 740) did encroach upon the round plaza partially, taking over about one half of the south-west quarter of the open space. A platform (an early ziyada) gave access to the structure from the main street, which was then set at a different level: see preliminary report of The Danish-Jordanian Islamic Jarash Project at http://miri.ku.dk/projekts/djijp/ (last accessed July 2012), with A. Walmsley pers. comm. The Yale team found a layer of debris on the plaza’s north-east side, which contained a hoard of unspecified 4th and 5th c. coins, under later “intrusive Arab walls” (by which the Early Islamic period is meant): C.H. Kraeling, “The south tetrapylon (Jerash)”, in Gerasa: City of the Decapolis ed. C.H. Kraeling (New Haven, Connecticut 1938) (103–15) 113. The current mission has identified that, in Late Antiquity, the oval plaza, along with other roads within the city, was covered by gravelly surfaces, with their paving not visible: A. Walmsley pers. comm., with I. Simpson, “Market buildings at Jarash: commercial transformation at the tetrakionion in the 6th to 9th centuries C.E.”, in Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham. Proceedings of the International Conference held at Damascus, 5–6 November 2006, edd. K. Bartl and A. a.-R. Moaz (Rahden 2008) (115–24) esp. fig. 3 (plan), with pp. 116–17, pointing out that the encroaching rooms of Umayyad date are in fact well-planned cellular shops, not chaotic structures. The deposition date for the hoard, within this surface, would likely have been no later than the first quarter of the 6th c., given the absence of reported 6th c. coins, suggesting that the gravel surfaces were in place by 500, rather than dating to after the Islamic conquest, as Kraeling’s use of the adjective ‘Arab’ here incorrectly implies. The tetrakionion was conceived at a time when the paving was still intact. Therefore, the gravel surfaces must post-date the construction of the tetrakionion, given by inscriptions on the monument, of ca. AD 292–305: see appendix F1 on tetrakionia.
Dating summary (decay as represented by gravel surfaces): range 293–500, midpoint 396.5, class Cs7 (TPQ, inscription), Cs3 (associative, finds), publication 2/3.
See also (K1a): 15ORI Antioch in Syria (Forum of Valens): The site seems to have been partially covered by a cemetery, from which ceramics, provisionally dated to the 5th to 7th c., have been recovered, giving a range of 400–700. A storage building with pithoi, built over the forum, the walls of which were cut down, has produced pottery of the same date. If the identification of this plaza as that of Valens is correct then we would expect this transition to be post-Justinianic given the monuments mentioned around the plaza, described by Malalas, active during his reign (e.g. horologion Malalas 13.30; macellum Malalas 9.5, 12.7).
See also (appendix S1): 15ORI Bostra (Bosra): Layers excavated immediately over the paving (reflecting occupation) contained unspecified pottery of the end of the 5th to the 6th c., whilst brown soil covering the paving (representing degradation) contained 7th c. material (which seems to be ARS, Jerash bowl, sigillata of Umayyad date, Gaza Ware and Grey Ware etc.). This places the degradation in the period 600–700.
See also (appendix S2): 10MAC Dyrrachium: The complex is dismantled from 570–600 and then graves of unverified 7th–8th c. date cover the plaza.
See also (appendix S3): 15ORI Scythopolis (‘Byzantine agora’): Occupation in the Umayyad period (AD 635–749), prior to a site-wide major earthquake, consists of the installation of a pottery production complex over the south-east portico and over part of the paving, whilst the cellular shops on the south-west side were rebuilt in a similar style to their predecessors. Thus, ‘messier use’ rather than abandonment seems to have taken place, unless there are two phases that have been accidentally telescoped into one during excavation.
See also (appendix U5): 15ORI Beroia (Aleppo): A church built near to or on the agora, perhaps in the second half of the 6th c., apparently before a mosque was constructed there in AD 715.
V5e Degradation and Disuse in the East, Undated within Late Antiquity
13ASI Nysa: Undated, unclassical structures were seen on the agora: L. Lavan site visit 2003, but nothing is mentioned in site reports that I have seen.
Dating summary: undated.
13ASI Pednellisos: It is likely that structures built over the agora were installed in the city’s final period of occupation, due to their good survival, albeit of flimsy construction: L. Vandeput, pers. comm., with L. Vandeput, “Pisidien Survey Projekt: erste Kampagne in Pednelissos”, Araştırma Sonuçları Toplantıları 20.1 (2002) (315–24) 317.
Dating summary: undated.
15ORI Abila (Decapolis): W.H. Mare, “The 1992 Season of excavation at Abila of the Decapolis”, ADAJ 38 (1994) (359–78) 363, describes a Roman plaza in the centre of the site being covered (and replaced) by a ‘Byzantine’ street (normally meaning 400–635 in this part of the world). No dating evidence is given in this or other reports that I have been able to access.
Dating summary: 400–635, midpoint 517.5, class 0, publication 1/3.