Dreya Mihaloew, Katie Rask, Marty Wells
ASCSA Corinth Excavations
North of Nezi
Session 2: April 27 – May 22, 2009

Session 2 excavations were carried out from April 27th to May 15th in the area north of the Nezi Field. The authors focused on the room between walls W55 on the north (N 1034.30), 6267 on the south (N 1026.97), 10086 on the east (E 282.15) and W54 on the west (E 273.95). Excavations in this part of the Byzantine house were previously carried out by William Berg in 1961(NB 229) under the directorship of Henry S. Robinson. In the first session of 2009, the excavations of Joanna Potenza and Ryan Boehm included this area. The 1961 trenches left the northern third of the room uneven and eroded. The central and southern portion of the room had been exposed by Potenza and Boehm down to Frankish and Late Byzantine levels.
Our objective was to continue the efforts of Potenza and Boehm in uncovering the 11th century levels for the purpose of public display. The director was Guy Sanders, the field director Alicia Carter, the pickman Thanasis Notis, the shovelman Tasos Kakouros and the barrowman Vasilis Kollias. The dry sieve was operated primarily by Sula Anastasopoulou, Kollias, Rask, Mihaloew, and Wells.

Frankish (1210-1458 C.E.)
Excavation suggests that the late-13th century saw a major reorganizing of the space in our area. One of the deposits that points to this activity is the fill inside Well 6288, which had originally been examined on May 13, 1961 (NB 229, p. 109) and designated Well OA-107 (coordinate designation) at that time. Berg excavated it for less than a meter and recorded no finds. He postulated that it belonged to the Turkish period and did not excavate deeper because of its narrowness. Our own investigation began by removing 0.45 m of backfill and debris before excavation. Initially we attempted to determine context changes based on differences in soil composition and inclusions; however, after approximately 2.0 meters of excavation, the Director advised that we should change contexts approximately every 0.30 m. This was our primary method of excavation, but we also changed contexts when stratigraphically necessary. All the material from the well was dry sieved with 7mm screens (although 3mm screens were used with context 6420 and thereafter); water flotation samples (15 L) were also collected from every context. Twenty contexts were removed in total, but the bottom of the well was not reached nor was the structure (6288) itself excavated.
The material removed from the well indicates three discernible dumping actions dating to the Frankish period, between 1270-1290 C.E. However, the character of these deposits suggests that the well was filled quite quickly, perhaps over a few days. The latest dumping layer (6286, 6291, 6297, 6360, 6361, 6365, 6368) fills the top 2.75 meters of the well. The contexts comprising the layer included a large amount of pottery, bone material, and various smaller finds such as iron nails, glass, and bronze objects. The proceeding (and underlying) dumping action revealed a dramatic decrease in the number of inclusions, with a very small amount of pottery and bone, and a significant increase in the ash and charcoal content of the soil (6369, 6371, 6374, 6378, 6383, 6386). Below and proceeding the ash deposit was another dumping action (6394, 6400, 6405, 6412, 6416, 6420, 6430); this deposit was characterized by soil with a high clay content and very little ash, extremely large amounts of pottery (with a high proportion of fine ware) and a very large amount of animal bone. In addition, we recovered small finds such as iron nails, a bone needle, spindle hooks and whorls, and glass. A total of 11 coins were found in this third deposit during this session.
Personal communication with Thanos Webb, the excavation’s zooarchaeologist, emphasized the distinct nature of the bone material that had been removed from the well. The preservation was very good, with little weathering and the presence of smaller and more fragile elements. There was an abundance of different anatomical elements, representing parts from the entire skeleton (e.g., the third phalanx, otherwise rare in the 2009 season). The surface modification of the bones was also distinctive, with the butchery marks on multiple elements being far more extensive than that on bones from areas outside the well. Finally, the species representation from the well was also conspicuous, ranging from common domesticates to large birds and fish, and with an age distribution ranging from fetal to mature. Additionally, large amounts of microfaunal remains and fish scales were collected from the dry sieve. These have yet to be analyzed.
The pottery found in the well dates to a twenty-year span (1270-1290), but the three dumping actions apparent amongst the contexts do not appear to be chronologically separate, despite their stratigraphic relationships. In addition, the presence of complete vessels suggests primary deposition, but the occurrence of incomplete and fragmentary body sherds also indicates the secondary deposition of pottery. This interpretation is probably supported by the bone material. The excellent preservation of delicate and small bones (e.g., of fish, cats, birds), as well of the articulation of some bones, can be indicative of primary deposition; on the other hand, a large amount of weathered and fragmentary bones suggests the secondary deposition of animal remains. The excavation of the well, including the structure, will continue in the third session. At this point there is not enough evidence to reconstruct the exact formation process of the well fills.
The second element of late-13th century reorganization involves the construction of architectural features. Potenza and Boehm removed the cobble foundations of two late 13th century piers framing a threshold on the southern side of the room (east: pier 5967, west: pier 5957). At the time these piers were constructed, two other piers (6148 and 6279), similar in size and plan, were situated approx. 3 m to the north and on axis with the southern pair. Both northern piers appear to have been constructed in the mid-12th century. The eastern pier (10088) and its foundation were in situ at the beginning of this session but have been excavated and removed. The western pier (6279) was robbed out in the second-half of the 13th century. This is apparent from two Frankish-era contexts, a pit (cut 6241, fill 6240) dating to the second half of the 13th century and the robbing trench of the pier (6254) which the pit disturbed, dating to the third-quarter of the 13th century. The removal of the pier foundation was possibly related to the construction of wall 10094, also dated, together with wall 10080, to the late 13th century. If our interpretation is correct, the space of the room was significantly altered in the late 13th century; according to the work of Potenza and Boehm, two N-S walls (10094, 10080) were added that split the original room, a drain (5938) was laid in the 3rd quarter of the century, several leveling fills were deposited, and new walls lined the southern extent of the room. Our excavation dated the removal of the north-western pier to the latter part of the 13th century as well. The filling and closing of the well in the years between 1270 and 1290 could have also been part of the change in use of the room.

Byzantine (10th through 12th centuries)
As mentioned, the second half of the 12th century saw the construction of the north-eastern (10088/6148) and north-western piers (6279, which may have supported an arch or an upper level. Potenza and Boehm likewise removed leveling fills which they dated to the mid-12th century. Our excavations recovered evidence for the earlier part of the century as well, represented by five fill deposits (6186, 6184, 6180, 6191, 6330) located in the south-eastern part of the room and laid within a long E-W cut (6199); in particular, several lenses of fill contained varied pottery of multiple periods and included large dumps of roof tiles (e.g., 6191: 21.4kg, 6186: 24.9kg). The mixed nature of the pottery deposit suggests that it was removed from another context before deposition in the room. There are at least two phases of 12th century activity: the fills date to the early part of the century while the construction of the northern piers dates to the latter half.
The earlier Byzantine levels of the room also suggest a period of extensive activity in 10th/11th centuries. A pebble surface is visible in the north-east portion of the room, south of the 1961 excavation trench and cut by the north-eastern pier foundation (6148). This surface was identified by Potenza and Boehm and left unexcavated. We removed four small deposits that remained on the top of the surface (6205, 6209, 6221, 6223), all dated to the 10th/11th century. The surface is yet to be excavated, and earlier surface phases are evident in the scarp created by the excavation of pier foundation 6148 to the east and the removal of fills inside cut 6224 to the south.
We have hypothesized that the central and southern parts of the room was disturbed by one large pit filled by successive layers of deposits. The 12th century fills discussed above cut into earlier deposits. We investigated a large circular pit (6224) with three fills (6225, 6229, 6334) dating to the 10th/11th centuries. These fills had previously been disturbed by other 10th/11th century activity, in particular the cutting of three probable post holes (6193/6194, 6195/6196, 6197/6198). These 3 post holes likely coordinate with a post hole (6217) to the north at the southern end of wall 10095 – the fill in two of the post holes date to the 10th/11th centuries (6196, 6217), while the other two are roughly dated to the Byzantine (6194) and Medieval (6198) period with no precise date. The post holes may have been meant to hold a light roof of some sort or some other wooden structure; their small size indicates that they could not have been supports for anything of considerable weight. We were unable to determine any relationship between the post holes and other floors and walls.
Our investigation of the stratigraphy in the southern portion of the room concluded with the removal of another lens of fill (6338) inside a suspected cut with the expectation that the nature and purpose of the cut would become more clear. This deposit was dated to c.1280 on the presence of one body sherd of Protomaiolica. This date is problematic as this material is cut by the pit 6224 and the 10th/11th fill inside (6234). We speculate that there was some contamination of this deposit as the fill was overdug at the western end. More excavation is necessary in this part of the room to confirm that the 13th century date is a mistake.
There is another option for this situation. It may be the case that the 13th century date for 6338 is true and that the fill deposits above it, though they contain 10th-12th century pottery, are actually depositional acts of the later 13th century with misleadingly low pottery dates. This situation will be investigated in the 3rd session.

Future work in this area should include:
1) the complete excavation of the well and the removal of its structure in keeping with the policy of open area excavation
2) the removal of any remaining intrusive deposits in the southern part of the room so as to establish the nature of the large suspected cut
3) investigation of the floor surfaces to the north-east in order to better understand the 10th/11th century activity
4) the four structures that were given numbers at the end of Session 2 (6421, 6422, 6424, 6426) should be further analyzed or excavated; establishing their relationship to other walls in the room will facilitate a better understanding of the room’s architectural and use phases.