North of Nezi (Green) Report End of Season 2009: Will Bruce (until 25 May), Scott Gallimore (until 1 June), Karl Goetze (from 25 May), Dan Leon (from 1 June)

The following summarizes the results of excavations during the entire season at Corinth 2009 in four areas north of Nezi Field previously excavated in the 1960s under the supervision of H.S. Robinson (NB 229, 230, 235, 253, and 264): the Byzantine courtyard, the room west of the courtyard, the room northwest of the courtyard, and the room north of the courtyard. These rooms were excavated in the first two sessions of 2009 by Scott Gallimore and Will Bruce; in the third session by Scott Gallimore, Karl Goetze, and Dan Leon.

Excavation in all four areas was supervised by Guy Sanders (director) and Alicia Carter (field director). Our pickman was Athanasios Sakellariou; our shovelman was Panos Stamatis; our barrowman was Sotiris Raftopoulos. Excavations were conducted in the southern area of the Byzantine courtyard (E264.40-E270.50; N1027.90-N1033.40), in the room west of the courtyard (E261.10-E264.60; N1030.80-1034.90), in the room northwest of the courtyard (E260.90-E264.60; N1034.85-N1038.75), and in the room north of the courtyard (E265.30-E270.50, N1035.60-1039.20). Periodically, overlapping stratigraphy or concern for phasing caused us to move from one room to the next. Our main goal in this area was to expose the walls and features of the Byzantine house for consolidation and presentation to the public. Our summary will be arranged chronologically by room.

COURTYARD

Frankish:
Session I began excavation in the courtyard when it was found that surface layers in the courtyard overlay context 6077 in the north room where excavation had been taking place. Session II began by removing wall 5508, which dates between the late 13th and early 14th centuries and was pedestalled in 2008 as permission to remove it had not yet been received, making it by far the latest feature in this immediate area. The construction of this wall limited access between the courtyard and the space in front of the rooms immediately to the south of the courtyard. The only point of access after the construction of wall 5508 was in the southwest corner of the courtyard. Session II excavation removed all remaining Frankish contexts from this area and all session III contexts can be dated from pottery and stratigraphic relationships to late Byzantine 10th-11th centuries.

Late Byzantine:
Session II excavation ended while excavating Late Byzantine floor surfaces and sub-floor leveling fills in the southern portion of the courtyard (south of removed wall 5508). Session III began excavation in this same area with the goal of finding a common floor surface that would link the area south of removed wall 5508 with the main courtyard area north of removed wall 5508.

Session III excavations in the southern part of the courtyard ended up concentrating on two areas: a series of floor and sub-floor leveling surfaces adjacent to all four sides of staircase 6296 which was pedestalled and left in situ and thus post dates all session III excavations, and the foundation trenches associated with wall 5783 (cuts 5795 and 6302) excavation of which was begun during the 2008 season but not finished.

Session III began by following a series of leveling fills and floor surfaces (contexts 6435, 6440, 6441) which were overlaid by context 6423 (excavated at the end of session II) and which ran successively north of each other until reaching cut 6302, the cut for the foundation trench for the west end of wall 5783. The pottery from the fill of the foundation trench dates wall 5783 to the late 11th century. Built pebble floor 6440 as well as its associated leveling fills 6423 and 6435, also date to the late 11th century and are the last contexts that can be dated to Late Byzantine period.

The southern courtyard area during the late Byzantine period saw the construction of pebble floor surface 6440 as well as wall 5783 and stairway 6296 which began closing off access to the courtyard from the south. Construction of Frankish wall 5508 effectively completed this process and so we can see a gradual evolution of this space from the southern portion of an open courtyard to what seems to be an interior corridor.


Middle Byzantine:

Session III next concentrated further on a series of patches, floor surfaces, and leveling fills in the area west, south and east of pedestalled staircase 6296 (contexts 6446, 6450, 6453, 6456, 6461, 6463). Successive removal of these contexts revealed a built floor made of pebble and tile that seemed to cover the whole area south of removed wall 5508 and which continues north into the courtyard proper (north of wall 5783 and removed wall 5508). We were unable to excavate this built floor surface due to the fact that it was cut by a robbing trench that robbed a section of wall 5519 and which would need to be excavated first. This area belonged to the yellow team however whose excavation had not yet reached this area. Although we could not excavate this surface, fills for two robbing trenches (6461 and 6463) that cut this floor surface help date the surface to late Byzantine, 10th-11th century.

Having finished excavation in the southern portion of the courtyard, we believe that we have revealed the latest surface to have spanned both areas: the courtyard and the room south of the courtyard, proving that in the 10th and 11th centuries, this was one unified space.

ROOM NORTH OF COURTYARD

Frankish:
Most of the Frankish levels of this room had been previously removed and were only encountered at the very beginning of excavation during session I. The only Frankish features encountered in this room, were three superimposed walls (5473, 5913, & 5914) which were removed according to a permit obtained this year. These three contexts were actually all components of the same wall from different phases: 5473 was the superstructure, with 5913 & 5914 as foundations. If any floor levels were associated with these walls, they must have been excavated during the 1960s, since the earliest floor from the 2008 season (5585) is late Byzantine based on the pottery from this context.

Middle-Late Byzantine:
Late Byzantine levels were encountered throughout the entire room, the latest being from the 12th century. The majority of the 12th century contexts found in the western half of this room consisted of a series of small leveling fills (5921, 5927, 5931, 5933, 5936, & 5962). A floor surface which may have been associated with these leveling fills was excavated in 2008 as context 5585. In the eastern half of the room, 12th century levels included a small pit (5926), several small leveling fills (5939, 5941, 5942, 5944, 5945, 5948, 5955, & 5958), and a threshold block (5865), which was removed. The floor surface with which this threshold could have been associated is context 5800, stretching from the doorway into the courtyard. The last 12th century context in this room was a leveling fill (5948) located in the northeast part of the room, the removal of which revealed an intact clay hearth (5975, 5976, cut 5977). The contents of this hearth were also water sieved, but nothing substantial was recovered. The only plausible floor surface, which could be associated with hearth 5975 would be context 5963. This floor was greatly truncated and did not come into contact with the hearth itself, but could conceivably agree stratigraphically based on their elevations.

The majority of the contexts encountered in the room north of the courtyard dated to the late 10th – 11th centuries. This seems to indicate that this was a period of intense activity in this room.
Session III excavations removed an extremely large dumped fill of large boulders (6516) which occupied the entire southern half of the room south of wall 6526 and west of wall 6016. Session I excavators who excavated down to this level before moving out of the north room interpreted this fill (6516) as composed of boulders from the superstructure of walls 6526 and 6016. Excavation of 6516 and other contexts (6532, 6550, 6554) around walls 6526 and 6016 serve to confirm this interpretation as all were dumped fills with significant inclusions of boulders. It is likely that walls 6526 and 6016 date much earlier than these dumped fills and that they were torn down specifically in order to level the surfaces in this room for the initial phase of middle Byzantine construction of this house.

Session III excavation also revealed a large pit (cut 6536) between (and disturbing) wall 6526 and wall 5562. Fill contexts for this pit include 6539, 6543, 6459 and 6460 - excavation of which (6460) was stopped when it was determined that the context went even deeper (and possibly into a cistern) and the pottery was late roman 5th – 6th century. This pit (cut 6536) however, took us down to foundation levels for walls 6016 and 6526 where they intersect on the north side of wall 6526 and the west side of wall 6016.

At this level there is evidence of a wall coarse of reused ashlar blocks on top of which wall 6016, made of mud and rubble, was built. Furthermore, a robbing trench was discovered above wall 6016 where it intersects with wall 5562 which served to rob stones from 6016 but which also uncovered the top of one of the reused ashlar blocks underneath. This ashlar block extends to the north into wall 5562 and either this same block, or another block on the northern course can be seen protruding from the north side of wall 5562. It was decided to differentiate this reused ashlar coarse from wall 6016 which sits on it and the new structure number is 6566. It should be noted that the tops of what looks to be more reused ashlar blocks can be seen at this same level under wall 6526. This evidence seems to indicate that these walls (6016 & 6526) are of middle or later Byzantine rubble and mud construction on top of late Roman walls of reused ashlar.

Further examination of wall 5562 west of the pit (context 6650) revealed the foundation trench for the south side of wall 5562 and the foundation trench for the east side of wall 6321. The foundation trench for wall 5562 truncated that for wall 6321 so it was excavated first. About .4m of this fill was excavated until we stopped due to concern that we would destabilize wall 5562. Pottery from this foundation trench was dated to Byzantine 10th – 11th century. Excavation of the foundation trench for 6321 was postponed while we searched for the continuation of the foundation trench for wall 5562 east of wall 6016 (context 6560=6554). We were unable to locate this trench, nor the one for wall 5990 on the east side of the room. At this point it was determined that we should move back into the northwest room and continue excavating in there. We did not have a chance to explore further the south side of wall 6526 and establish whether it is indeed part of the same construction as wall 10111 and how such a wall would relate to the rest of the house.


Room NW of the courtyard

Frankish:

The latest feature in this area was a well that had been left intact by previous excavations. The well is located at the eastern end of the room, and cuts many earlier deposits. We did not excavate all the way to the bottom of the well, but removed fill to an elevation of 83.80. The well widens towards its base and seems to have been built in two phases, a main construction of tile and plaster (6493=6638) with a superstructure of small cobbles and plaster at the top of the well (5876). A perplexing aspect of this well’s construction appeared at a level much below the top of the foundation and embedded within contexts dated firmly to the middle Byzantine period. What appears to be a cut for a foundation trench traveling around the northern half of the well (6641) was overlaid by contexts that must be much earlier than this well. It is possible that some sort of slump or collapse accounts for the cut, rather than deliberate human action, and this event may be related to the slumping that was visible at several levels in the area immediately north of the well, similar to that discussed by Lima, Webb, and Kolb in the notes for context 5558.

We also excavated leveling fills, a wall foundation, and fill inside a tile-built sub-floor storage feature, all dating from the Frankish period.
The latest of these features was a cobbled wall foundation (6100), the superstructure of which was removed last year as context 5604. Beneath this context was a firmly packed soil surface.
We performed a cleaning inside a pit excavated last year as context 5644 to determine whether last year’s excavation had reached the bottom of this context. Within pit 5644 was a small fill (6097) and a robbing pit (6103, cut 6106) associated with wall 10111. We continued excavating pit 5644 as context 6115 (cut 6116), and it became clear that this was the fill of a subterranean tile-built storage feature with a depth of ca. 1.10 meters. None of the actual structure of this feature was discovered, but the fill and the cut made its identification certain. A comparandum lies in the room north of this one (context 5504), excavated in 2008. Within the fill of 6115 were found a well-preserved late-Roman Ionic capital, a stone mortar, and a large assemblage of Frankish pottery. As for reconstructing the use life of this tile-built storage feature, our excavations up to this point can only inform us that this tile-built storage feature was out of use by the late Byzantine period since it was truncated by wall 5725 which is currently dated to the Late Byzantine period.
Removal of the storage feature fill (6115) also gave us insight into the construction history of two walls (10111 & its N-S jog 6130), which predate its construction. A portion of wall 6130 was uncovered during the removal of 6099 & 6107, but was not identified as a wall until 6115 was excavated. We hypothesized that wall 10111 is a continuation of the E-W wall (6526) uncovered in the room north of the courtyard.
In addition to pit 6115, most of the Frankish contexts excavated in this room were fills (6107, 6108, 6110).

Late Byzantine:

In the NW corner of the room, directly beneath a layer of Frankish fill, we discovered what appears to be a large rubbish dump, comprised of contexts 6607 and 6612, which together reached a depth of 1.68m. However, we did not finish excavating the dump because we realized, when we reached the bottom of the surrounding walls (5562, 5142, 6130, 10111), that it was later cut by all four of those walls, and thus earlier, that is, the dumped fills ran under these walls. This dump yielded a large number of cooking sherds showing evidence of heavy use, as well as a great deal of material associated with cooking such as egg shells, fish scales, and animal bones that had been chopped or sawn. The volume of such material suggested commercial, rather than private, use of the area. Sanders has tentatively suggested that this part of the Byzantine house may have been used as a tavern in this period, and that the dump was an associated rubbish heap. Such usage would accord well with the hypothesis offered by Gallimore and Bruce in the summary for Session II concerning a large deposit of white-ware kettles resting on a floor (6129). The date for this deposit based on these kettles (1100 ± 10) provides the best chronological marker for the construction history of the levels excavated in this room which must be Late Byzantine or later. Moreover, In the southern portion of this room, several leveling fills of late-Byzantine date (6151, 6153, 6154) had been deposited for the construction of a floor, likely excavated in the 1960s since no floor is noted in this area from the 2008 excavations. These fills covered a small semicircular cut (6156), the fill of which (6155) was serving to backfill the tile-built storage feature discussed above. We discovered a floor in this part of the room (6496) which was a part of the same floor upon which the white-ware deposit sat (6145), and which also yielded a large number of white-ware sherds that had presumably been stamped into the floor from above. This floor was firmly dated by pottery to the same period (1100 ± 10) and was later cut by the aforementioned tile-built storage feature (6519), which itself was cut by wall 5725, suggesting that the NW room may originally have continued further south and joined with the W room, though such a connection.

Middle Byzantine:

On the E side of the room we excavated a series of floors that had been disrupted by the Frankish well. One of them showed evidence of a repair patch (context 6588), and all of them overlaid a robbing trench that ran along the E boundary of the room, to the S of the well. This trench was filled with a mixed deposit that initially made the identification of a single trench quite difficult. Another difficulty was that wall 6375, to which the trench grants access, seems not to run to the full northern extremity of this room. Sanders suggested that to the S end the wall had only been partially robbed out, but to the N end the wall had been completely robbed out by some later action to accommodate construction of the well. Such an action is difficult to reconstruct however, since the trench was overlaid by middle Byzantine floors, which would have to have been removed by the Franks as they built their well, if the robbing was indeed associated with that construction project. Alternitavely, wall 6375 did not continue north of wall 10111, and so the robbing trench would not be expected to be found north of wall 10111. The date of this robbing action presents a problem, as the stratigraphy and pottery suggest Middle Byzantine in the northern sections, but three sherds at the southern end (fill 6349) are of Late Byzantine date, and Sanders has given a very firm date of 1150 ± 10. It is possible that these sherds are contaminants, but a stronger likelihood is that further excavation will force the revision of the dates of some surrounding contexts, many of which are less than specific, eg Byzantine 10th-11th century NPD.


Room W of the Byzantine courtyard

Frankish:

A series of floors and sub-floor leveling fills were excavated in this space. The fills (6159 + cut 6161, 6162) seem to be leveling fills for an unidentified floor, perhaps excavated in the 1960s, since no upper floor surface was noted during the 2008 excavations. Revealed by removal of these fills were two isolated patches of floor, one (6163) in the northeast corner, and the other (6165) in the northwest corner. These two patches perhaps belonged to the same floor surface, but were kept separate in the Harris matrix, since their elevations do not correspond. Associated with this floor were twelve leveling fills: (6167, 6174, 6176, 6179=6182=6185, 6189, 6192, 6203, 6178, 6207, 6214, 6218, & 6236). These floors and fills were laid up against wall 5725 because they overlay its unexcavated foundation trench. Context 6165 was cut by the foundation trench (5720) for wall 5762 in the west. Thus, this floor surface, if the same, postdates wall 5725, but predates wall 5762.
The removal of one of these fills (6203) revealed two distinct courses of wall 6228 (previously labeled 50kj in the 1960s records). The top courses (structure 6206) were much more crudely built and had no associated foundation trench. A mid-late 12th century sherd embedded in the soil matrix of upper courses made it clear that it was a later construction and thus we removed it on April 30th. During the 12th century the inhabitants may have added to the height of the wall on account of the rising floor level created by the addition of fills and floors.

The floor level revealed by the removal of all of the aforementioned twelve fills was context 6237. We were able to excavate one fill below this floor (6239), but we ran into difficulty because robbing trench (6350) located immediately east of this room truncated the surface beneath fill 6239 (6465), and thus we had to shift our focus temporarily to the southern half of the courtyard.

In a pit in this room, located in the southern part and truncated by the northeast corner of the Ottoman house, we discovered a small extension of the cut and unexcavated fill, which we excavated as context 6214 (cut 6215). This pit was excavated in the 1960s, but we have not yet identified which of the 1960s notebooks refers to it.

Sakellariou believed he could discern the cut and fill of the foundation trench of wall 5725 in the east scarp of “Bothros 9” (NB 235, p.19), which disturbed most of it. “Bothros 9” appears to be of Frankish date, based on Guy Sanders’ examination of the lot pottery (Lot 837). This foundation trench cut has as yet not been revealed by the further removal of deposits in the area (see Middle Byzantine, below).

Late Byzantine:

In this area we excavated a series of floors that had been disturbed by later actions. Floor 6477 had a post-hole dug into it which would be consistent with some sort of roof support. It was also cut in the south by what appears to be a pit filled with large debris, although this pit was itself disturbed by a later pit, which was excavated in the 1960s, and so its purpose remains unclear.

Middle Byzantine:

The series of floors and leveling fills from the Middle Byzantine period were excavated in this room, one of which (6484) shows evidence of a post-hole similar to the one discovered in Late Byzantine levels. This post-hole was probably also used for a roof support. In this period the room seems to have been used for cooking, since a hearth (6653) was cut into the fill at the southern end of the room (6658=6664). The same fill was cut dramatically short of the southern wall of the room for a purpose that remains unclear. It was in this area (fill 6658) that a fragment of the Roman sima of the South Stoa was discovered, namely a terracotta head of a female divinity. The earliest material excavated in this room was a floor that was broken in many places (6670=6668=6672) that does not show evidence of cooking, so it seems that cooking only took place in this area for a limited time during the Middle Byzantine period. The removal of floor 6668 revealed a slump that may be related to the foundation trench for wall 5725, but no firm cut has been discerned. Since wall 5725 cuts a Late Byzantine tile-built foundation feature (see above), and thus must be a Late Byzantine or later wall, it seems unlikely that a foundation trench will be found below this level and that the foundations of this wall were constructed flush with the foundation trench cut.


Conclusion:

All four of the areas excavated by Team Green during the 2009 season are components of the Byzantine house under investigation in the area north of Nezi Field. They represent four independent, but interconnected spaces which appear to have undergone substantial modification during their use. Our excavations this season have clarified many aspects of the construction history in this area of the house. It appears now that the rooms immediately northwest and west of the courtyard were divided during the Byzantine period by wall 5725, and may have originally been one large space. The relationship between the courtyard and the room immediately to the west has also been clarified by the excavation of robbing trench 6350 for wall 6375. When wall 6375 was removed, probably during the Late Byzantine Period (early-mid 12th century), this opened the courtyard to the west and made obsolete threshold 6320. The removal of this wall seems contemporary with the pier foundations 6359 and 6318, which represent the construction of a support system for a roof over of part of the courtyard, thus diminishing the open-air area of the courtyard. Access to the courtyard was reduced during the late 13th or early 14th century with the construction of wall 5508, which left only an entrance at the southwest as a point of access leading toward threshold 6320 in the northwest corner of the courtyard.

The high-volume cooking activity apparent in the NW area (cut by walls 5562, 5742, 5725, and 6321) suggests that for at least part of its history the house may have been used as a tavern, which may also account for the frequent renovations. The true nature of the cooking activity may be revealed by the further demolition of late walls such as 10111 and 6130. The Frankish well would ideally be removed as well, though there may be some practical difficulties owing to the nature of the shaft, which expands at its lower levels. The NW room’s relationship to the W room has become much clearer, and the removal of wall 5725 should make it easier to treat the two as a single space.

With respect to the room immediately north of the courtyard, there remains at least one deposit of leveling fill atop the earlier E-W wall in this space that dates from the period when the first phase of the house had gone out of use and the floor levels were raised for the construction of the second phase of use of the house. This leveling fill may reveal floors associated with the first phase of use of the house. However, attention must be paid in keeping excavation in phase between the room north of the courtyard, the courtyard itself and the rooms NE and E of the courtyard so as not to artificially break up earlier features/deposits associated with the first phase of use of the house.