Corinth Report: Kosmopoulos Trenches excavations 2020, by Belza, Anna (September 28 - November 20)
Collection:   Corinth
Type:   Report
Name:   Kosmopoulos Trenches excavations 2020, by Belza, Anna (September 28 - November 20)
Title:   Kosmopoulos Material From the National Archaeological Museum at Athens Returned to Ancient Corinth Museum
Area:   Kosmopoulos Trenches
Site:   Corinth
City:   Ancient Corinth
Country:   Greece
Anna Belza, PhD Candidate University of Cincinnati ASCSA Corinth Museum Project Volunteer
Fall 2020–Spring 2021
Project: Alice Leslie Walker Kosmopoulos, repatriation of Prehistoric material from the National Archaeological Museum in Athens


Alice Leslie Walker Kosmopoulos was a student of the ASCSA 1909–1914, and associated with the School until 1937. She was assigned the study and publication of the pottery from the Corinth excavations (ca. 1896–1935); later the scope was narrowed to the pre- Byzantine pottery, and eventually to only the Prehistoric period material. The material she included in her study were from her own excavations at Corinth (1911, 1914, 1920, 1930, 1935)
and those of other excavators (1904, 1905, 1908, 1916, 1926, 1931, 1932).

Kosmopoulos conducted her study at Corinth before relocating to Athens ca. 1935. Her reasons for moving were twofold: her poor health which was worsened by conditions at Corinth (e.g., dampness, mosquitos—she had previously contracted malaria at Corinth); her expulsion from the Corinth excavations due to her falling out with the ASCSA. Kosmopoulos writes about her interactions with the school in the preface to her published work: The Prehistoric Pottery of Corinth (1948). In sum, conflict between Kosmopoulos and the ASCSA regarded her poor/nonexistent publication record. Kosmopoulos responded by relocating some Prehistoric material from Corinth to Athens.
The Prehistoric pottery from Corinth was stored at the National Archaeological Museum (NAM) at Athens in order to facilitate Kosmopoulos’s study and publication process. When the ASCSA severed ties with Kosmospoulous (ca. 1937) they demanded the material be returned to Corinth. Some material was returned to Corinth and is referred to in the Corinth storage system as the Kosmopoulos series or K- series. A large quantity of pottery remained at the NAM following Kosmopoulos’s death in 1954.
Kosmopoulos published one volume on Prehistoric Corinth in 1948. The introductory volume provides basic insight into her ceramic classes and chronological scheme (see Appendix 1). She did not publish all the material that was removed from Corinth to Athens. Attempts were made to return the material to the Corinth Museum1 (viz., Lavezzi in the 1970s–1980s). Robert Bridges visited the NAM in the 1980s and did a basic inventory of the Corinth material. In September 2020 the material was returned to Corinth. The quantity and quality of the material was unknown.
My museum project involved: the unboxing and processing of the Prehistoric Corinth material returned from the NAM; separating the material into lots; and entering all the material into the Corinth records. The prime objective was to process material quickly in order to learn what Kosmopoulos had taken from Corinth and glean evidence of Prehistoric activity otherwise unknown. The quick processing benefited positively PhD candidates Jeffrey Banks (University of Cincinnati) and Katie Fine (Florida State University) who are writing dissertations about Early Bronze Age and Neolithic Corinth, respectively.2
After all the material was sorted, it became clear that it was possible to lot the pottery (more on this below, Phases 2 and 3). I also transcribed the Kosmopoulos label notebooks (Appendix 2) which were given to Ioulia Tzonou to eventually be incorporated into and assigned Corinth Notebook numbers. We do not have Komospoulos’ excavation notebooks from Corinth (the ASCSA archives have her Halae notebooks). Their exact whereabouts is unknown. At one point, decedents from her husband’s side of the family living in Peiraeus attempted to sell a trunk that belonged to her to Henry Robinson. Robinson declined to purchase the trunk blindly (i.e., without knowing the contents) at their high price: without knowing whether the notebooks were there, the trunk and its contents would have been a waste. Jeffrey Banks has attempted to reach out to the family members with no success as of yet (May 2021). Banks believes that the Corinth notebooks went to California with Kosmopoulos in the late 1930s. Kosmopoulos finalized the publication of her book in California and was then in the process of a planned second and third volume. It is almost certain that she had them with her as she continued her work in California because the second volume was meant to be a large presentation of the prehistoric material from Corinth. Problematically, the Halae books did make it to the ASCSA archives which was meant
to form part of the third volume of her study, but that body of material is much smaller and there are indications she was finished with it.

Phase 1: Processing the Kosmopoulos Material Returned to Corinth from the National Archaeological Museum, Athens, in 2020
Kosmopoulos used cardboard shoeboxes to store and transport her material. The material placed inside the shoeboxes was grouped in smaller packages of paper tyropita bags, reused envelopes, and, in some cases, loose in the box. Material occasionally was found grouped in clear plastic bags, likely a solution by the NAM to replace old paper bags that had decayed. The storage shoeboxes were transported to the basement of the Corinth Museum in thirty-three wooden storage trays. Processing occurred in the basement of the Corinth Museum.3
The system for processing material involved using iDig on an iPad and recording various information in Microsoft Excel on a Corinth project laptop. James Herbst and Manolis Papadakis set up a context labeled Kosmopoulos in the iDig database. Every “shoebox” was photographed: before opening to record all markings on the exterior of the box; when opened with the contents left in situ, showing the storage of the bags and interior markings; and unboxed with objects strewed. Shoeboxes often had writing on their exterior in one or more crayon colors. The writing was often illegible, written over/crossed out and remarked multiple times, and/or contained a series of undeciphered abbreviations. All but one shoebox were discarded after being photographed and all such marking recorded. In large part, these marking could not be deciphered. It is clear that some of the marking referred to the one time contents; but with the multiple reuses of the boxes and multiple packing and re-strewing of her material in Corinth and Athens over a twenty year period, these marking did not seem to correlate in any meaningful or useful way for what was stored within when they were opened in 2020.
The contents of the shoeboxes, mainly ceramics, were strewn on three tables in the Corinth Museum basement for processing. Contents from each shoebox were kept and photographed together so that any given object could possibly be associated/reassociated to the markings on the box if they are ever deciphered (i.e., everything from one shoebox was laid out together on a table). Sherds stored together in bags or envelopes were laid out atop the bags from which they came. The bags and envelopes also often had illegible and/or abbreviated handwriting in crayon. A few times, typewritten text was used instead. In some instances, these markings were clearly a count of sherds of various types (her classes or the colloquial classes of pottery at that time) stored within: e.g., “5–Urf[irnish], 2–B[lack]B[burnished], 3–Myc[enaean]”, etc.
It was often not possible to discern why Kosmopoulos separated sherds into individual bags, if not by diagnostic features or grouping of decorations. For example, all material of a single class is not conveniently grouped together nor are groups associated based upon their excavation context. The divisions of bags seem to reflect the process by which Kosmopoulos read the objects, recorded them, and stored them, probably working in small batches of pottery because of space issues at the NAM and in light of the considerable amount of material she was working with.
Kosmopoulos seems to have generally followed Wace and Blegen’s classification system for the Early Bronze Age and Wace and Thompson’s system for the Neolithic, though she did not use the same terminology/abbreviations as far as we can tell from the markings on bags and boxes (e.g., she often prefers German terms, likely from her work with Dörpfeld and study of the prehistorics at Leukas). Sherds contained inconsistent markings (discussed further in Phase 2 and Phase 3) that reveal various information: their find spot, their depth, and year excavated. The markings played a large part in our ability to re-lot the material and for Banks to recreate the original contexts for his dissertation (described in Phase 2 below).
Objects of note were removed to receive inventory numbers (CP or MF), discussed below; all those not selected for inventory numbers were stored together with other objects from the same bag (i.e., the Kosmopoulos bagging system was the organizing principle of recording individual “units” in the initial sorting and identifying of material). Some, such as Neolithic gray wares, were often boxed together even if sherds came from different envelopes in one shoebox. There was no clear reason to distinguish the Neolithic graywares and there were no sherd markings or envelope markings that would give a reason why they should not be combined and the storage units pared down. The original envelope context was recorded in the processing photos.
Once material had been processed, they were stored in Corinth Excavation cardboard boxes (open top) which were placed into wooden storage trays. Each box received two numbers: a “K-NAM” number and “box” number. K-NAM numbers represent the shoebox in which items were found: these numbers were not from the original boxes, but were assigned based on the order in which they were processed; they simply help identify all the material that was originally boxed together in the NAM (i.e., to retain an association of the material with the marking on the original shoeboxes). Box numbers identify subdivisions of storage within the shoe boxes: most often, this was simply the paper bags or envelopes within which items were stored. Again, the box numbers only reflect the order in which the material was processed, they were not derived from information on these bags. Box identifies items grouped together in Kosmopoulos bags; K- NAM identifies the shoebox in which larger groups of these bag/box objects were stored.
The individual cardboard boxes were then placed in a wooden tray in the basement of the museum. During the sorting and recording of these items, many objects were assigned inventory numbers: 530 objects were assigned Corinth Pottery (CP) numbers; 45 were assigned Miscellaneous Find (MF) numbers. Their original Box and K-NAM numbers were recorded with these inventoried pieces, but the objects were disassociated from the box/tray system described above.4 Inventoried objects were recorded in iDig as “Objects” and photographed individually. Later, each inventoried object was fully measured and described according to the Corinth Excavations recording system. A running list of the CP and MF numbers were printed out and left with the crates. Inventoried objects were set aside in their own trays in the Museum basement for conservation and photography to eventually process (as of June 2021, the objects have not been conserved or photographed). Eventually these will enter the study collection. A Neolithic expert in particular should go through the objects and vet whether all of these specimen are worth retaining as CPs/MFs, particularly in light of the greater number of objects selected.
K-NAM and box numbers and CP and MF numbers were recorded in three excel sheets. At the beginning of the processing system, before the use of iDig (about a 1-week period), we
were not recording shoebox/K-NAM numbers, as we were still attempting to discern Kosmopoulos’ recording and storage system. In order to record what processed objects/tyropita bags were found together in one shoebox, we recorded that information as “Packed in K-NAM Shoebox with K-NAM Museum Box #”. Using context numbers in iDig rectified this problem; however, we continued to record boxes found together and this was superseded by the K-NAM number system which was retroactively applied to all of the individual box units that had been sorted prior to the advent of this system.
For example, this is the format used to record processed material:

Other fields used in excel are:

Bag or card info to signify if there was any written information found with pottery on their bag or card:
Contents: sherd count, general chronology, shape, fabric.
CP assigned to objects: CP number(s) given to something from that box. MF assigned to objects: MF number(s) given to something from that box.
Other notes: includes comparanda or publication information in cases where these objects were published in Kosmopoulos’ book (book, page number, etc.), notes to Jeffrey Banks, Katie Fine, or Ioulia Tzonou about specific items from the box that may be of interest to their EBA, Neolithic, and Mycenaean studies, respectively.
Recorded in iDig: whether or not it was recorded in iDig (Yes = yes; Blank = no). Notations on sherds: markings in pen or pencil that were legible, originally made by
Kosmopoulos to preserve excavation context information.

In total 173 KNAM shoeboxes were processed into 267 cardboard boxes. Two boxes were found in the NAM material that need to be returned to Athens. The first is an orange box with pottery ranging in date from EBA–Classical, obsidian, and a loom weight. Pottery find spots were recorded on the sherds (e.g., Thera). A notecard was found inside the box stating that the material was seized from the German Archaeological Institute (DAI). It seems to be a study collection and seems to have nothing to do with Kosmopoulos or Corinth, returned to Corinth with the Kosmopoulos material by accident. The second box contained numerous tags from various sites (not Corinth), all placed within a foam mold for a bronze spearhead or knife. Banks believes that the latter box might have been the commercial packaging for a knife that was used to hold a bronze dagger found by Komopoulos at Corinth, which she dated to the Middle Neolithic (it was not, but almost certainly EBA); the whereabouts of the dagger are unknown, and it is likely to have eroded away or remains in Athens.

Phase 2: Establishing Original Contexts for the Kosmopoulos Material from the National Archaeological Museum
Kosmopoulos abbreviated original excavation context information (e.g., trench and/or year, depths) in pencil or pen on many sherds, almost certainly whenever she removed material from its original context-tagged pottery storage tin.5 Sherds determined to be insignificant were grouped in tins, labeled, and stored at Corinth in the Old Museum.6 All this material was originally taken to the NAM by Kosmopoulos but returned when ASCSA demanded the return of all Corinth material and severed ties with Kosmopoulos in ca. 1937. “Insignificant” material was returned, while the “significant” material remained at the NAM (i.e., highly diagnostic objects that seemingly would have been published in the Kosmopoulos planned—but never finished— volumes on prehistoric Corinth).
Jeffrey Banks and I sorted sherds back into their original contexts. Pottery was separated into trays based on the markings on them that designated their findspots. The EBA and Neolithic pottery were kept separate within their context units to facilitate future study. The process took place in the Museum basement and courtyard.
Two additional columns were added to the KNAM excel sheet to keep track of markings on sherds and where items were being combined/lotted: Kosmopoulos Area Notations on Sherds and Re-lotted. The former recorded markings found on sherds (e.g., “E35”). If notations
were illegible or difficult to distinguish, they were returned to their box and placed in a tray for future revisiting (highlighted in the excel sheet in orange so that we could return and reprocess these after an initial sorting). The latter column (“Re-lotted”) recorded whether items were re- lotted (yes or no) and, if so, into what trays they were combined (e.g., E35, 2TH, 11 Heer 7, etc.). In some cases, all the material from boxes were inventoried (i.e., received CP numbers) and thus did not get lotted (e.g., see table below—“No context pottery to sort”). In many cases, all material from a box was lotted by context, and that box number no longer exists as a discreet storage unit, other than as a recording unit for objects’ original location.

Banks partook in the process in order to better understand where the Early Bronze Age material was found and to see if it was possible to rectify stratigraphy based on elevation markings on some sherds. He was able to use the sherd markings and Kosmopoulos trench system and depths to reconstruct a number of contexts across the site and combine this information with her publication, various excavators notebooks, and archival material to get a full understanding of what most of the sherd markings mean.

Kosmopoulos Series in New Apotheke and Old Museum: Preparing it for Lotting with the K- NAM material
After processing all of the K-NAM material (i.e., the Kosmopoulos material that was returned to Corinth in 2020), Banks and I went to the ASCSA Apotheke7 to examine the “Kosmopoulos Series”8 material had never left Corinth, or which was returned to Corinth by Kosmopoulos in the 1930s. This was around Christmas break (Dec. 25, 2020–Jan 15, 2021) when the Italian conservation team vacated the facility for the holidays. At the end of this period, when the conservators returned, this “Kosmopoulos series” material was moved to the Old Museum so we could continue our work.9
The Kosmopoulos series material stored in the New Apotheke was sorted and examined previously by John Lavezzi and Katie Fine. Lavezzi had sorted the EBA and Neolithic material based on chronological periods and distinct wares (e.g., red slipped rims) to facilitate an eventual attempt to combine the NAM material and look for joins. Katie Fine sorted four trays of the material Lavezzi had not managed to sort while a regular member as museum project. Fine’s sorting grouped material based on features of sherds: (e.g., Prehistoric–Roman rims or bases) regardless of chronology or context. Both these sorting methods were no longer relevant in light of our greater understanding of the original excavation contexts which had become the primary lotting principle of the Kosmopoulos material.
Banks and I applied the same sorting technique described above to the material in the Kosmopoulos Series: sherds were separated into boxes based on the notations about original excavations. In total we sorted through twenty trays. The contents included: ceramics, figurine fragments, stone tools (various), and shells. Four trays were unsorted/unstudied material ranging in date from the Neolithic–Roman periods.
In January 2021, Banks and I began to work in the Old Museum courtyard with (ca. 48) trays of Kosmopoulos Series material.10 Mostly, this material lacked individual sherd markings, and the impression is that this was the “insignificant material” Kosmopoulos left behind in Corinth or sent back. Based on Banks’ understanding of the history of the Kosmopoulos material and its various storage and papsing processes it received while in Corinth, these were almost certainly stored in tins that distinguished original context and depth; at some point this information was lost when the material was combined into trays and the original storage tin units lost. Some tags were included in boxes within trays, making it possible to glean, at times, where some material originated from, although almost all of these tags identified that the sherds within had come from more than one context.
Material that could be assigned to a specific context were combined with the proper excavation context/lot units that had been assigned for the KNAM material and the Kosmopoulos Series material from the New Apotheke.11

Phase 3: Assigning Lot Numbers to Context Pottery

Once all of the K-NAM material was sorted by context, Ioulia Tzontou, Jeff Banks and I agreed that lot numbers could be assigned to the pottery based on original excavation units (for the most part, these refer to identifiable/spatially known trenches). The lotting could not have taken place if sherds had not been marked with excavation data (e.g., trench abbreviation, depth). Banks provides full detail about the lots and contexts in the study for his dissertation and is in the process of generating lot descriptions. The lotting process is ongoing as of June 2021: the final quantities of material that cannot be assigned to a specific context will have to be considered (e.g., combine all Kosmopoulos unidentified location material to a single lot, lots based upon possible locations, toss some material, etc.).

Topographical Reconstruction of Prehistoric Habitation at Corinth

The K-NAM material attests a larger spatial and chronological use of the site than known previously. The quantity of material returned doubled the amount of known Prehistoric ceramics found in excavations. For more information the topographical reconstruction of the site with deposit information, see Banks’ dissertation.
Chronological Implications

Weinberg’s publication of Neolithic–EBA material from his excavations remain an important source for understanding Prehistoric activity at Corinth.12 The K-NAM material offers a more nuanced understanding of chronological periods because of the quantity and quality of material and the fact that they derive from deposits across the site. See above and Banks’ dissertation for a thorough discussion of the relevance of the Kosmopoulos material.
It is unclear whether Kosmopoulos saved all the Prehistoric material from her excavations. It seems likely when one considers the amount of Final Neolithic grayware body sherds she saved. It remains possible that Kosmopoulos intended to papse material at a later date but never finished with the material or had the time to do so. This is especially true of the later material excavated in the 1930’s were the extreme volumes of material and particularly the inclusion of what would normally be termed “insignificant” sherds suggested a near to 100% retention of excavated material, at least until they had been studied.
Below is a rough count of the pottery from the K-NAM processing. It is meant to give an idea of quantities representing chronological periods. The number will surely change after specialists complete their studies. The numbers represent the Kosmopoulos material returned to Corinth from the NAM in 2020 (i.e., they do not include the Kosmopoulos material that had already been in Corinth since the later 1930’s).
EN: ~16 sherds, including 1 mendable variegated bowl (CP 3967) MN: ~70 sherds
LN: ~3,536 sherds EH: ~1,433 sherds
MH: ~2 sherds (CP 3970: Gray Minyan goblet; CP 3977:1 possible Standard matt painted figure 8 around handle)
LH: ~10 LH III (CP 3974–3976)

A few of the LH sherds were marked with “Zyg”, or “Zyg dump”. From Kosmopoulos’ publication, these almost certainly refer to a pile of pottery that had been dumped outside of the Old Museum: it included Bronze Age Zygouries and Neolithic Lechaeion Road East material and excavation unit tags and seems to have been thrown out after Blegen and Hill fell out with ASCSA and were no longer working at Corinth. Kosmopoulos recovered the material. These sherds were placed in the Zygouries study collection drawers in a small bag with a printed explanation included.

1 For the sake of posterity: Corinth Museum refers to what is often referred to as the “New Museum”; more clearly, this is the contemporary Museum function currently (2020–2021). There are plans to build a new New Museum, so this may cause confusion in the future.
2 For a detailed biographical/archival analysis of Kosmopoulos and her work on Corinth and Prehistoric Greece, see Banks’s dissertation (forthecoming).
3 The basement provided poor light, and in many cases information and notations gleamed from Kosmopoulos were more apparent when viewed in the sunlight at, e.g., the ASCSA Apotheke on Asklepius Street. An additional reading of all the sherds in a more suitable location may reveal additional details of Kosmopoulos’ work and methodology, particularly since her notebooks are missing and the seriation of her excavation units could only be recreated by Banks based on the depth markings on sherds.
4 Ioulia Tzontou (Assistant Director) selected sherds and objects to receive CP and MF numbers, particularly for the Neolithic. Jeffrey Banks selected Early Helladic sherds to receive CP numbers based on their relevance of specifically for inclusion in his study of the EH period for his dissertation and later publication.
5 Kosmopoulos 1948, p. 8, fn. 19.
6 For the sake of posterity: since there is a new museum being planned, “Old Museum” might refer to one of two structures. Old Museum is the original/first Museum, which currently functions solely as a storage space and makes due as a study space, located on the south side of Apollo Street, just west of the village plateia, along the northern edge of the archaeological site—west of the exit gate and east of the Roman North Market.
7 For the sake of posterity: this storage facility if currently (2020–2021) referred to as the New Apotheke. There is currently plans to create another Apotheke and either also create a fresco lab or retain the New Apotheke as a frescolab as it is currently functioning as such. For sake of clarity, this apotheke is on the north side of Asklepius street, east of Cheliotomylos, northwest of the main archaeological site and current Museum.
8 The Kosmopoulos Series is a term used to refer to the Kosmopoulos material stored in Corinth in tins (later in trays) that were assigned “K” numbers for storage recording (K-1, K-2, etc.)
9 See fn. 5 for more on the Old Museum.
10 See fn. 5 for more on the Old Museum.
11 See fn. 6 for more on the New Apotheke.
12 Weinberg, S. S. 1937. “Remains from Prehistoric Corinth,” Hesperia 6, pp. 487–524.