Repository location

https://doi.org/10.7910/DVN/IU1XFO

Context

This data was produced as part of a larger project attempting to better understand the distribution of cuneiform objects in colleges and universities around the United States (US) and its territories. It serves as a complement to the ArcGIS web map, “Where is the Cuneiform?”¹ (Mohr, 2020) which displays the specific locations on college and university campuses where one may find these objects held or on display. The dataset used for “Where is the Cuneiform?” overlaps with, but is ultimately separate from, this dataset. Cuneiform objects from early excavations of ancient Iraqi sites were widely distributed among collectors through both legal and illegal means, leading to varying levels of information regarding these objects and their original context. Providing the locations of these collections and the accompanying bibliography represents a crucial first step in a clearer understanding of their provenance.

Steps

The dataset was produced from data gathered via the process of creating the accompanying ArcGIS map and via the bibliographic information provided by the Cuneiform Digital Library Initiative (CDLI)². In the process of running Google searches with targeted key terms to identify cuneiform collections at US colleges and universities, I was able to find scholarly editions of the objects in some of these collections published on the institution’s website. The vast majority of the data came from the CDLI. Each entry in the CDLI has an optional field for publication information, where one might find a unique publication dedicated to that specific collection, or a publication in which the object appears as a duplicate of a well-known text. However, the CDLI bibliographic data is incomplete and required much restructuring in order to create the full citation. For example, the CDLI offers “D.I. Owen in Fs Lipinski” as the citation for the collection at Oregon State University. Using Brown University and University of Chicago library catalogues, I was able to construct the full citation: “Owen, D. I. (1995). Amorites and the location of BAD. In K. Van Lerberghe & A. Schoors (Eds.), Immigration and emigration within the ancient Near East: Festschrift E. Lipinski (pp. 213–219). Leuven: Uitgeverji Peeters en Departement Orientalistik.” Each bibliographic entry in the dataset is presented in the APA citation style. Further information such as author, title, and publication year are separated into separate columns for ease of organizing the data based on that information. All information was then organized into a CSV file organized by the name of the college or university holding the collection.

Sampling strategy

Because the focus of the wider project is meant to be on those collections that are less well-known or smaller in size (typically 5–20 objects), there are several collections that are included in the ArcGIS map that do not appear in the bibliography for two primary reasons. First, while many cuneiform collections are known around the US, not all of them have had their contents published. Second, some of these collections are both incredibly large and very well-known among Assyriologists and those outside the field. In particular, the following collections were not included in this dataset: Yale University, The University of Chicago, Harvard University Museum of the Ancient Near East, and the University of Pennsylvania. Not only would the inclusion of these collections be unfeasible for the project’s scale, they are large enough and well-known enough that their publication records are much more easily accessed. Readers can expect to find the following institutions cited in the bibliography in Table 1.

Quality control

The decision was made to cite each source using APA-style citation practices in order to keep the dataset consistent. Once the dataset had been created, it was cleaned via OpenRefine. As each publication was identified, it was entered into a spreadsheet according to the institution where each collection is housed. Cleaning via OpenRefine was necessary in order to ensure that each bibliographic entry was paired to the correct author and the correct collection. Using the clustering function in OpenRefine, I was able to see which entries the computer considered to be the same. Inconsistencies in this organization were then corrected to ensure that any user could reliably sort the dataset by institution, author, year, and publication title.

Dataset creators

Sara Mohr (Brown University)

Roles: conceptualization, data curation, investigation, methodology, writing (original draft & review and editing)

Language – English

License – CC0

Repository name – JOHD Dataverse Repository

Publication date – 2021-01-24