This paper builds on research we have carried out in the context of the pre-oil Gulf region with the 2000-page “geographic and statistical” section of the Gazetteer of the Persian Gulf, Oman and Central Arabia, a compendium of British imperial intelligence compiled by John Gordon Lorimer in the early twentieth century. Our paper explores the possibility of combining toponymic information from Lorimer’s Gazetteer with that of other historical sources covering overlapping geographical areas, including Ottoman material on southern Iraq and the Arabian Peninsula, British gazetteers of India, and material produced by other imperial interests in the region (Omani, Saudi, Qajar, French, German, etc.) Lorimer’s geographical and statistical dictionary, which represents two-fifths of the entire Gazetteer, exhibits a tension between geographical description and statistical catalog of elements of agriculture, animals and human built space. The text’s toponyms operate on a number of levels of specificity that do not necessarily correspond to the British administrative vision of the region, ranging from salt flats to tracts of desert to human settlements. Likewise, the over 800 entries in the geographical dictionary contain sharply varying levels of specificity that nevertheless follow certain identifiable patterns: human settlements designated as towns, for example, largely include descriptions of built environment, human geography, animals, agriculture, markets, and defences, some of which are organized tabularly. That the level of detail also varies by region within the Gazetteer points to regions where the British had a long historical presence or a strong network of local agents. We will discuss how the structure of place in Lorimer’s Gazetteer does not reflect a hierarchical vision of the region, but rather an expansive and intrusive imperial desire for granular information about spaces in which British sovereignty was always highly contested and legibility was assumed to be inadequate. It considers the implications of choosing any one of these sources, particularly Lorimer, as a “spine” gazetteer, against a comparison of the structural assumptions of various sources. Instead, the paper discusses the first stages of data creation in the Open Gulf Project where a diversity of sources reflect overlapping imperial interests and geographies and divergent levels and scales of imperial legibility.

The proliferation of digital humanities and spatial history methods in Ottoman and Middle East studies is currently impeded by the absence of a digital gazetteer for the Ottoman world. Such a resource would include historically-attested variations of a toponym, trace its administrative dependency in time, its geographic coordinates (which may also vary in time), alongside a wide variety of relevant information. Such a resource would not only be be the backbone of any digital humanities project, but would also standardize and integrate existing regional digital gazetteers that are built for the purposes of individual digital humanities projects. The paper focuses on the methodological and technical challenges of trying to build such an essential infrastructure for digital humanities scholars. It will describe the previous efforts taken towards that end, and will then focus on the methodology developed in the context of the ‘Ottoman Recogito’ project that was awarded a Pelagios Commons Resource Development Grant in 2018. Engaging a group of graduate and undergraduate students along the lines of what may be called ‘pedagogical mini crowd-sourcing’, the project used the Recogito semantic annotation tool (https://recogito.pelagios.org/) in order to extract toponymic data from volume 12 of the “Important Affairs Registers” (Mühimme defterleri). The first step was to identify Ottoman toponyms that are currently available in large online gazetteers (GeoNames, Pleiades, etc.), and some 1,000 place names were identified in the source processed here. The second stage of the project entails identifying Ottoman toponyms that are not found in existing gazetteers and will require further research in order to find their geographic coordinates as well as other relevant data. These two pools will be the basis and the foundation of an Ottoman Gazetteer that will have Unique Resource Identifiers (URIs) and will be freely available to the online community.

Ottoman space is not flat. The toponyms that appear in Ottoman records are nodes in a hierarchical administrative vision of provinces, subprovinces, districts, and townships. A complete gazetteer of Ottoman space must represent not just the geographic coordinates associated with place names but also their place in the imperial schema. A gazetteer of this kind will be a resource not just for the reconciliation of historic place names but for more comprehensive study of the Ottoman administrative gaze. It would permit scholars more accurately to estimate, for example, how many kazas the empire contained in any given year, and to situate assertions about the past in a clearer context. A tool like this would be especially useful in studies employing quantitative evidence. A gazetteer of this kind requires an ontology, which is a controlled classification schema and a set of rules specifying the relationship between elements of that schema. The best practice in linked data is to employ an existing vocabulary; geonames offers a hierarchy of administrative districts that generally corresponds to the needs of an Ottoman gazetteer. Many of the puzzles of Ottoman history are products of its administrative exceptions, however. The nature of a vilayat-? mümtaze cannot clearly be captured by geonames or other generic schemas. Ottomanists need to develop their own ontology. This paper describes a basic approach to this task, using linked data tools. Representing the Ottoman administrative hierarchy is a relatively straightforward matter of specifying the rule and a limited set of exceptions. The greater challenge is representing change over time in the gazetteer. This is a problem that I have not seen a clear path to tackle, and I hope in this paper to propose a method to periodize this representation.

“KIMA - Towards an Open Hebrew Gazetteer” set the grounds for a comprehensive, dynamic and interoperable database of historical place names in languages written in the Hebrew Script. Each entry in this database consists of preferred forms of a toponym (both in Hebrew script and in its English normalized form), Alternate Hebrew script names and their transcriptions, together with their extant historical attestations, a calculated earliest use of each variant, and geographical coordinates where available. An attestation based gazetteer, which aspires to offer not only name variants and coordinates, but also information on any known use of these toponyms through time, can serve not only as tool or reference but as also as a corpus which can be read, albeit with caution. In the first phases of the project, the largest source type for this information in the first stage of the project was library catalogs, which document the place names written on the title page of books. Mapping this corpus reveals patterns in the history of Hebrew letter printing, and it would be wrong to read it as an indicative map of Jewish existence. In order to have a more refined look manuscripts and correspondences should also be incorporated; Finally, an altogether different space will be revealed when toponyms which appear within the texts are included, this time showing not where the Hebrew letter was used but rather what was the conceived, even imagined geography of the world that was described in it. Each source type carries its biases and thus a gazetteer building is a constant amendment of biases in previous sources with the addition of new sources. The current phase of the project, carried in the framework of the project DiJeSt (DIgitizing JEwish STudies), which aims at addressing another bias in the sources: while the overwhelming majority of attestations in the corpus are Hebrew, other languages that have made use in the Hebrew letters are under-represented, and among them Ladino and Judeo Arabic. Assembling a database of attestations for Judeo-Arabic toponyms not only augments and corrects the bias created by Hebrew language gazetteer, it also enables exploring continuities and differences with the Arabic toponymic practices. In order to facilitate such exploration, we incorporate a transliterating tool. The paper will present the Judeo Arabic resulting toponym database and the geo-temporal patterns it displays.