Ceramic review

Laerke Recht – March 2026

Ceramic review procedure

To make the analysis of the ceramic sherds of each digital unit book available in the UGR, a detailed review is carried out for each unit book.

This review ensures the consistency and quality of the presented data. In some cases, a more extensive conversion of shape codes is required in order to fit the format of the developed typology and methodology.

In broad terms, the review includes the following steps (not all unit books require all steps, and not always in the order noted here):

Inputting scanned handwritten ceramic analysis sheets, usually into Excel or a similar open format
Collating and consolidating all files related to a specific unit in two complete 'cumulative' files: one for shape sherds (diagnostic), and one for body sherds
Checking for duplicate entries (several entries with the same q-lot and p number); solving all identified duplicates
Checking that all ceramic codes are valid (including for ware, shape, decoration etc); solving any incorrect codes
Checking which q-lots and features have been analysed (and which might still be missing)
Checking that all q-lot and features listed in the Excel sheets correlate correctly: solving any incorrect correlations
Checking that all drawings and images match the data from the cumulative files; solving any occurring problems
When the shape cumulative sheet is entirely 'clean', convert all shape codes to the updated typology, following the established protocol
Reviewing, updating and reformatting all assemblage pages
Creating the overview of ceramics of each unit book, for the lefthand side of each book - see e.g. A15, J3, J5.

In the individual unit books, the results of the review can be found in three main pages. Taking J6 as an example:

On the righthand side, under Constituents, qp
On the righthand side, under Constituents, assemblages
On the lefthand side, Under Typology/Objects/ceramics

Typical errors

While many of the errors and challenges identified during this procedure are individual and require individual solutions, there are some that commonly occur and might be expected after 25 years of excavation and a system that is continuously updated and improved based on new knowledge, including:

Earlier formats and types of recording. This includes both digital formats and/or software that is no longer accessible and systems of recording. For any one unit, a unique situation may have resulted in an adaptation of a suitable means of recording. In many cases, this can be solved through technological solutions to conversion of older files, and through the memory of members of the team who participated in the fieldwork.
Typos: the switching or misreading/mislabelling of numbers, especially when data are transferred from handwritten to digital format.
Discrepancies in the correlation between q-lot and feature. This is typically due to a typo or a specific procedure in the field.
Differences in the analysis of the ceramicist and illustrator (e.g. for rim diameter or orientation of a sherd).

Transparency and history of observations

In order to maintain full transparency of the actions taken during the review process, corrections made to a specific sherd/q-lot/feature are noted in the pertinent record.

Any sherds removed from the record due to inconsistent data are kept on a separate list, along with the reason for their removal in each case. Our philosophy is to retain as many entries as possible, and to solve all discrepancies. However, if there is still substantial doubt about the correct context (q-lot/feature) or details of a sherd or set of sherds, that sherd or set of sherds has been removed.

Not all q-lots in every unit book has associated analysed ceramics. In most cases, this is because no ceramics was found in that particular q-lot. More rarely, the dataset for that q-lot was not available. Therefore, a list of analysed sherds (shape and body sherds) is also created for each unit book in the process of the review of the ceramics.

A learning process

The problems that caused the errors were due to both the sheer quantity of the material (in line with our policy to analyze not only every vessel but also every single sherd excavated) and by the changes in the recording system. This serves as a lesson for the future: we feel confident that when excavations are resumed, the new records will be in excellent conditions at the end of each excavation season.

People

The review of the ceramics of each unit book is supervised by Marilyn Kelly-Buccellati and Laerke Recht, and has benefitted from the assistance of Caitlin Chaves Yates (A16, J1, J2, J7), Jessica Scaciga, Lukas Gran (A15), Tobias Welz (A15), Brigid Clark (A10, J4) and Emilio Semidei (A6, A7, A12, A20, J7).