Macellum: August 6-10, 2006

Excavation of the Macellum continued for a fifth week under the direction of Julian Richard (K.U.Leuven), Mustafa Kiremitçi (Dokuz Eylül Universitesi, Izmir), and Christine Beckers (K.U.Leuven). After the excavation of two shops--Rooms 3 (4.60 by 3.46 m) and 4 (3.30 by 4.77 m)--during the last weeks (see Macellum: July 30-August 3, 2006), the team focused on the area in front, i.e. to the east of them, where last week the first remains of the Macellum's western portico were exposed (see Macellum: July 30-August 3, 2006). It became clear that during its second phase of use Room 4 had a doorway (1.27 m wide; 0.46 m deep and 1.67 m preserved height) made of recycled column shafts recarved as doorposts and, to the right of it, a large window (1.47 m wide; 1.45 m preserved height) with similar recycled posts and a sill supported by a brick wall (0.80 m high) on the outside. The "stepped" bank or counter inside the room (see Macellum: July 16-27, 2006) is just behind this window. The question arises whether or not it functioned as a real counter during phase 2 of the complex, then, after the abandonment of the Room during the sixth century (see below), was transformed into a kind of stair at the level of the destruction layer containing the contemporaneous water supply outside (see below). By then, that had become the walking level so that people still could enter the room for dumping material inside or carrying out other activities there, as in fact the actual door had been blocked up to the same level as the window sill.

An in-depth study of the findspots of coins and other datable artifacts offered some better insights into the complicated stratigraphy. In fact, an original construction date for Room 4 in the late second century A.D. (most probably under the emperor Commodus) was not contradicted by the coins from the floor substrate below the largely removed pavement slabs. With the exception of two fourth-century A.D. coins (the emperors Valens and Valentinianus), they all dated between the reign of August and the later second century. The whole layer must have been mixed and contaminated though, during a later intervention, since it ceramic evidence mainly pointed to the sixth century, more precisely to the period ca. A.D. 560-620. One of the sherds even belonged to a vessel found in the erosion/destruction layer of Room 3. This clearly shows that despite the fact that so many rich artifacts were left inside both Rooms 4 and 3, people had removed not only some of their paving, but also searched through their floor substrates. On the other hand, the above-mentioned concentration of rich artifacts above the floor level (see Macellum: July 30-August 3, 2006) contained many coins dating from the late fourth to the first quarter of the fifth century A.D., whereas those of a layer dumped above them (locus 70), possessed a concentration of coins ranging in date from the mid-fourth until the mid-fifth century A.D., the most recent one even being a follies minted under the emperors Heraclius and Heraclius Constantine (A.D. 610-640). The latter date fitted well with that of the pottery from the same locus, which could be attributed to the first half of the seventh century, or more precisely to the years A.D. 580/90-640. However, restudying of the pottery of locus 24 is advisable here, as the ceramologists believed both layers to be stratigraphically related, whereas the trench supervisors made a clear distinction and accepted the existence of an abandonment phase between the rich deposit (locus 24) and the dump above it (locus 70). The picture in Room 3 and even in the portico itself (see below) was rather similar. The floor substrate of the room at the back contained mainly coins dated to the fourth and fifth centuries A.D., whereas the dumped layer above the floor provided numismatic evidence dated, with the exception of a Hellenistic Ephesian silver coin (see Find of the Week), exclusively between the very end of the fourth century and the reign of Anastasius (A.D. 491-518). Yet, although the excavators assumed an abandonment phase here as well between the rich deposit and the dump above it, related with traces of a walking level, the evidence is less clear than in Room 4. Locus 48 in this room, corresponding with a destruction and erosion layer, full of roof material (especially tiles), which seem to have fallen down over a long period of time, contained coins from the later fourth century, but also from the reigns of Justinus II (A.D. 565-573) and Phokas (A.D. 603-610), whereas except for some residual material, most of its ceramics could be attributed to the years ca. A.D. 560-620.

During the first half of the week, the crane removed more architectural blocks. On that occasion, three more inscribed frieze blocks from the west portico architrave were found, belonging to another part of the inscribed entablature found last week (see Macellum: July 30-August 3, 2006). Beside providing the usual information about P. Aelius Akulas, the founder of the Macellum complex, they confirm that the name of the emperor to whose victory the monument had been dedicated was almost totally erased, over a length of ca. 1.50 m. This means that on the west portico much more than just the first, the family and one of the surnames (supposedly Commodus) of the emperor's titles were erased. In view of the fact that on the architraves (almost certainly from the south portico) copied by K. Lanckoronski in 1884-1886, the beginning of the dedication could be read as "Because of the victory of the Lord Emperor Caesar Marcus [then missing: Aurelius] C[ommodus Antoninus and], a text occupying no less than three architraves, which is the same as the erased part on our preserved pieces, there seems to be nothing abnormal here. Erasing the isolated name "Antoneinos," which on its own does not make any sense, may just have been forgotten.

A comparison of all architrave parts recovered thus far with those copied by K. Lanckoronski, who assumed the presence of three almost identical building inscriptions along the West, South, and East, suggests that as he wrote himself that most of the inscriptions he saw were spread along the slope towards the Odeion, most of his text (his version I, which is almost complete) must have come from the supposedly open south portico. From these blocks we recovered, mainly from the lower part of the slope, two of this eleven architrave parts, numbers 6 and 10 of which the first one was broken in three pieces (see Macellum: July 30-August 3, 2006). Block 8 of his third building inscription, in reality, also belonged to the first building inscription, the southern one. On the other hand his assumption to have copied parts of at least three almost identical building texts is correct, although there must have been a fourth one, still completely buried on the north side, where the entrance was located, as well. Without any explanation, Lanckoronski had completed the building costs of the Macellum on a missing part of his southern building inscriptions as being 13,000 denarii. However, most of our fragments come from the western portico, almost completely buried at the time of Lanckoronski's visit. There the part with the real costs is preserved and clearly mentions that the whole construction has cost P. Aelius Akulas a total of 50,000 denarii. During the Antonine period these silver coins had a weight of nearly 2.3 gr. At the time of Nero, its weight was still 3.41 gr. Halfway in between the reign of the latter (A.D. 54-68) and the construction of the Macellum ca. 180-191 (see Macellum: July 30-August 3, 2006), the rebuilding of the peristasis (colonnade) of the Temple of Apollo Klarios at Sagalassos and the covering of its internal walls with marble veneer, had cost in A.D. 103-104, besides the not specified amount of money contributed from their own fortune by T. Flavius Collega and his wife, another 10,000 denarii. If one assumes that the private contribution of Collega did not surpass what he collected from others, as he did not mention the amount of it, one could imagine a cost around 20,000 denarii for the whole building transformation of the Apollo Temple. Even taking into account the increasing inflation and decreasing value of the denarius between Trajan (A.D. 98-117) and Commodus, the building costs of the Macellum must have far surpassed the Apollo Temple repairs three generations before.

Although, there were clearly minor differences in the texts along the different sides, the dedication on the west portico can now be more or less reconstructed as follows (missing characters are indicated between square brackets): [Because of the victory of the Lord Emperor Caesar Marcus Aurelius Commodus] Antoninus and for eternity dedicated to the most brilliant and sweetest of fatherlands, the first city of Pisidia, friend and ally of the Romans, [P. Aelius Antiochos Aquila/Akulas, son of Neon, grandson of Rhodon, great-grandson of Konon son of Konon], next to the other generosities which he bestowed, has also given as high priest of the emperors 50,000 denarii for the construction of the Macellum.

Study of the building blocks gives us a better understanding of the layout of the 4.27 m deep portico surrounding the central courtyard. They comprise column pedestals, fragmentary column shafts (estimated total height: ca. 3.60 m) made of the expensive white-veined dark gray kaplan postu marble from the Docimian quarries (near Afyon), a white Corinthian capital, inscribed architraves below a fluted frieze (a so-called Pfeifenfries), and large-sized Corinthian cornices. This abundant architectural evidence permits a complete reconstitution of the portico's original layout. At the time of the emperor's Diocletian's A.D. 301 prize edict to stop inflation, one cubic foot of Docimian (pavonazetto) cost 200 denarii.

Toward the end of the week, we reached the level of the portico's floor. In the meantime, we discovered four encroachment walls subdividing the portico in smaller units, of which the extremities were lost inside the late water-supply system (see below). Most of the original pavement, except for the exterior edges, probably made of slabs, had apparently been removed at a later date. The whole portico had been filled with a ca. 0.60 m thick destruction layer characterized by a great amount of ceramic and faunal remains. Only a few glass and metal artifacts were observed in it, together with some coins, the most recent of which could be dated to the reign of Anastasius (A.D. 491-498). The ceramic remains, however, point to a date from possibly the late sixth century up to ca A.D. 625.

A late water-supply system was found on top of this destruction layer and runs throughout the portico in a north-south direction. It is made of mortared rubble covered with well-aligned, rectangular rubble stones, among which there was a re-used inscribed fragment. The direction from which the water-supply system comes, most probably suggests that this is an extension of the very late water-supply system discovered in the 1990s along the southern edge of the Upper Agora and built on top of the latter's slab pavement. As it continued southward, it is possible that this is the same system identified along the blocked east "VIP entrance" to the Odeion below it (see Odeion: July 10-13, 2006), which perhaps continued on top of the collapsed west facade of the Roman Baths along the east side of the Lower Agora. In recent years, we had not even excluded that the latter possibly carried drinking water to the mid-Byzantine kastron (fortified village of the tenth-twelfth century inside the shrine of Hadrian and Antoninus Pius; see Apollo Klarios Sanctuary: July 30-August 3, 2006). Yet, the fill dumped between this water channel and the entrance to Room 1, contained exclusively coins issued between A.D. 393 and 429, although the material can clearly have been removed from elsewhere to dump here. On the contrary, the ceramic evidence from this fill, besides the presence of a small though clear quantity of residual (older) material, in majority points to a date in the sixth century, perhaps its second half up to ca. A.D. 625. The datable sherds from the foundation trench of the water supply system, however, could be placed between A.D. 480-525, perhaps up to A.D. 550, rather suggesting an early sixth century date. In any case, when the west portico collapsed, some of its columns broke while hitting the somewhat raised water-supply system. Parts of the entablature (e.g. the upper part of Room 1's south pillar) also cover the latter. So we know that the Macellum's western portico still stood at least partially upright when this water system was built (see Macellum: July 30-August 3, 2006).

All of this brings us back as to the date of the collapse of the Macellum's shops and portico's. The excavations of this and last year clearly showed that the original construction almost certainly had been built during the reign of Commodus (A.D. 180-191) according to the emperor's name, but that it had been rebuilt probably twice (evidenced by columns recycled as door and window posts and the fact that the latter do no longer occupy their original position). At some point, while Rooms 3 and 4 contained expensive luxury goods, a catastrophe seems to have been occurred, which caused the building to be suddenly abandoned, leaving a lot of valuables behind, although the construction remained, at least largely, standing.

The trench supervisors noticed also signs of an abandonment phase, clearly visible in Room 4, and less prominent but suggested by a walking level in Room 3 as well. Afterward, the remains inside the shops and the portico were covered with dumped material (with sherds from a same vessel originating from a dump and a floor substrate being recovered in different rooms!). This shows that at the time of the dumping, those responsible for it, must have been aware that the rooms still contained precious goods and also looked for them. At the moment, it seems that the dumping activities occurred mainly during the second half of the sixth and the adjoining part of the seventh century, whereby Room 3 seems even to have been searched as deep as its floor substrate. Around the middle of the sixth century at the latest, a water-supply system seems to have been arranged inside the still-standing portico, which however seems to have been gradually decayed with part of the roof falling down, from the later sixth century A.D. onward continuing into the earlier decades of the seventh century, before most likely the earthquake of the 640s completely toppled the structure, which fell on top of the well-built water-supply system, causing its columns to break apart. This also means that, unless repaired later, this water channel is not related to the water-supply of the mid-Byzantine kastron in the shrine of Hadrian and Antoninus Pius (see Apollo Klarios Shrine: July 30-August 3, 2006). The cause of the abandonment of the Macellum is still uncertain. Yet, the date of the erosion and destruction material found inside seems too late to link it to the earthquake ca. A.D. 500. On the other hand the fact that the whole complex seems to have been abandoned rather suddenly, with a lot of precious things left behind, then temporarily abandoned and eventually used as a dump, collapsing very gradually from the middle of the sixth century A.D. onward, might suggest that the plague of A.D. 642 had possibly eliminated the owners or tenants of the shops or even the whole complex, which was given up for some time before gradually degrading. As a place, which normally should also have contained food shops, this is easily understandable.

The west portico is bordered on the east by two steps (each 0.30 m high), still in situ. The column pedestals from last week were found almost lying in their original position on the upper step. Below this short staircase--which marked the limit between the central courtyard and the portico--the original pavement of the former was exposed. It consists of ca. 0.70 m long, 0.45 m wide and 0.25 m thick limestone slabs aligned longitudinally. Just below the steps, a row of slabs placed 0.06 m higher than the rest of the courtyard's pavement, contained a curved gutter running along the staircase (W.: ca. 0.30 m; Depth: ca. 4 cm). It was most probably collecting rainwater both from the portico's sloping tile roof (although the lions' heads on the cornices were not pierced) and water flowing on the pavement.

Toward the end of the week, the excavated area was extended northward, in two sectors, in order to find the rest of the portico/central courtyard. There as well, the same kind of architectural remains--two more in situ column pedestals, column shafts, and cornices--as well as another well-preserved section of the floor pavement were found. All structures found in situ in the portico still date from the original building phase of the Macellum, during the second half of the second century A.D. However, some traces of repairs (such as clamps holes cut onto the inscription of one architrave) indicate that parts of the portico had undergone subsequent damage and repairs, possibly already as the result of the earthquake around A.D. 500. During the next weeks, we hope to expose more of the central courtyard's paving, and to extend the investigated area toward the central tholos of the Macellum. We should also place back a certain number of the architectural fragments found this week and the week before: at least four pedestals can be placed into their original position, together with one column (reconstituted from three fragments that will be glued with epoxy) and its Corinthian capital found last week. The purpose will be to give an impression of the original layout of the Macellum, from the central tholos to the shops at the back with the portico in between. In this way, we should obtain a cross-section of the Macellum, from the early Byzantine shops at the back to the second-century A.D. architectural structures of the courtyard.