The Record of Objects

The Records of Objects are the primary artifact documentation produced by the Michigan excavation, comprising eleven bound volumes covering finds from Karanis, Soknopaiou Nesos, and Terenouthis. Each volume is titled with its archaeological contexts and contains typed pages that were almost certainly copied from handwritten field notes. The volumes document many of the identifiable artefacts discovered in a building or unit, though the data are far from total — smaller, less-easily identified objects were almost certainly thrown into the spoil heaps.

The volumes contain 68,438 entries, some of which include multiple finds. As was customary for the time, the finds were divided between the sponsoring organization and the Egyptian government: the Egyptian Antiquities Service distributed 2,462 artefacts to various museums, while 38,011 field entries representing at least 46,415 artefacts entered the Kelsey Museum collections, and 2,603 ostraca, papyri, and wax tablets were catalogued in the University of Michigan Papyrology Collection. More than 24,989 artefacts were left in the field. The recording system itself evolved across the seasons.

B224B*

Placed at the top of the page, this header indicates the precise room with which the following objects were associated. The “official” designation, which was written on the artifacts, would also list the year, 1930, abbreviated as 30. The first B indicates the excavator identified “occupation layer."

The excavtors believed they could identify multiple periods of occupation at the site. The top or “A layer” represented the last period before abandonment, which they suggested was the fifth century. The “B Layer” was earlier, probably in the third century. 224 is the building number, while B represents the room. The asterisk indicates material found in a particular space, most likely under the floor level.

Artifact Designations

In the 1924 season, artifacts were assigned sequential capital letters (A–Z), with ceramics given lowercase letters, and objects of the same type grouped under a single designation. From 1925–26, Roman numeral suffixes were introduced to extend the system; if more than 26 artifacts were found in a particular locus, they would be numbered AI, BI, and so on. Peterson made further refinements from 1926–27 onward.

Annotations

This indicates that several beads were recovered in the unit. The annotation in red shows that six were discovered, given numbers 77022, 78113, 78248, 76159, 76757, and 76954. After each bead, a short description is provided, sometimes with a coded typology. For example, the second bead is made of faience, of type IV.A.1.

Two Frags, Glass Inlay

The first number, 6358, is the Kelsey Museum Accession Number. In the field, these two glass fragments were inscribed with the field number, 30- (the year) B224 (the building) B (the room), * (beneath floor level) -B (the artifact number).

The second annotation, 5 of 7.2521, is a reference to a division photograph, taken before objects were divided between the University of Michigan and the Egyptian government. Shown with other glass fragments as number 5, the glass inlay is visible in the second row, two down from the number 55.

Objects Left in the Field

These objects, for which there are no further annotations, were likely left in the field. The Michigan team recorded the find, but it was not deemed sufficiently important to catalog or retain.

Ostracon

Cataloged as part of the University of Michigan Papyrus collection (view here), this ostracon, or inscribed potsherd, records the delivery of chaff to Karanis by an individual named Oualas.

The translation reads “On Pachon 27. Of Karanis. (Has paid) Oualas, 1 basket (sargane) of chaff."

Pottery Lamp

The lamp, like the beads (artifact A) and the ceramics shown below, were cataloged by typology number, following the best practices as suggested by the pre-eminent Egyptologist of the day, Flinders Petrie. The Karanis excavation employed a lamp typology that was probably borrowed in from the publications of the site of Ehnasya. Further work on this typology by the Karanis Housing Project team will provide additional insights into the meaning of this type listing.

Ceramic Vessel, Type 55

Ceramics were listed with lower case letters, and, after the first season, followed the pattern discussed above, where Roman numerals were appended to the lower case letters for artifact numbers above 26. Here, the ceramic vessel uses the ceramic typology as shorthand for the documentation.

This vessel is type 55. The remainder of the code provides a basic description of vessel, which is a bowl on a base ring with a moulded rim. The sides are carinated, with the carination at or above the middle of the vessel.

The sketch of the vessel helps to better identify it. The annotation to the side, NTH, indicates that the artifact was “Not Taken Home” or left in the field. More information on the ceramic typology can be found here.

Marginal notes in the current bound volumes list museum accession numbers, and dates were added over time for papyri, ostraca, and coins as subsequent publications revealed additional information — a testament to the volumes' ongoing role as living documents rather than static field records.

Yet, the volumes present significant interpretive challenges for researchers. The recording system used letters, special characters such as the asterisk, and Roman numerals — an array of field designations that makes it difficult to transfer the records into a modern database or to conduct effective searches or analyses.

To address this, the team developed a new naming system that remains traceable to the original while being fully computer-readable. Alphabetic characters were converted into numerical equivalents, and the asterisk — a common wildcard character in computing that would cause problems in a database — was replaced with a caret (^). For example, field designations such as B224B* would resolve into a predictable new format: B_0224-002.^. The level, if present, is listed first, followed by an underscore. Then, the building number, resolved to 4 digits, separated by the room, translated into a number. In this case, 224 becomes 0224, and B, the second letter, becomes 002. Finds, for the artifact number, are separated by an additional underscore. Ceramic objects received an additional distinction: because vessels had been recorded separately from other artifacts using a lowercase letter suffix, the new system places a "c" before the ceramic number, so that a find like B224B*-a becomes B_0224-002.^_c001. This approach means that anyone familiar with the original field records or earlier publications can easily reconstruct the old number from the new one, preserving continuity with decades of prior scholarship. Artifact descriptions were standardized using terms from the Getty Art and Architecture Thesaurus . Ceramic finds received their own sub-system, translating the excavators' proprietary pottery code into expanded, human-readable descriptions.

All cleaned artifact records were loaded into a PostgreSQL relational database aligned with established metadata standards — specifically ArchaeoCore (NYU Institute of Fine Arts) and the Archaeological Resource Cataloging System (Michigan State University). Each find record was then enriched with links to external open-data sources, including the Kelsey Museum's Artifacts Database and papyri.info . The Karanis Housing Project is currently building a searchable, stand-alone web-based database that can be queried for information about finds, findspots, and archaeological contexts. For now, the web map offers similar capabilities.