In the 19th century, religious and political interest in Jerusalem led to a prolific production of maps depicting the city. Notably, none of the retrieved maps were created by local cartographers; early examples were often produced by individual travelers without formal cartographic training, and later by Western nations and organizations. Factors such as limited expertise, restricted local knowledge, and difficult accessibility resulted in significant inventions, omissions, and distortions. At the same time, the strong interest in the subject and the difficulty of conducting new surveys encouraged the widespread copying of existing maps, which led to a proliferation of these inaccuracies.
This phenomenon is indeed a limitation for accessing the urban history of the city, but it can also provide a rich source of previously unstudied historical knowledge, not only for tracing the origins of individual errors but also for correlating them to reveal underlying copying relationships. By interpreting these errors, we aim to situate the maps within their inherent subjectivity. In this thesis, 200 maps of Jerusalem from 1810 to 1925, the largest dataset analyzed to date for this period, are systematically examined to study these misrepresentations and to use them as sources of historical information. By drawing a parallel with the fields of stemmatology and textual criticism, this thesis treats distortions and misrepresentations as "divergences" that can be analyzed locally to trace the history of an area's representation and globally to derive genealogical relationships between maps.
Focusing on three key aspects (topographic content, textual elements, and planimetric accuracy), the study introduces tools and methods to extract and interpret these divergences. The planimetric layer, which encodes the distances and bearings between mapped features, is selected for computational analysis because of its fundamental differences from the other two: while place names and topographic content are best handled by human expertise in such a diverse dataset, planimetric distortion patterns need to be transformed into interpretable data through quantitative analysis. Novel tools are developed specifically to automatically georeference part of the maps and to subsequently measure and compare planimetric distortions across the full dataset.
The results are used to automate the reconstruction of genealogical relationships among the maps, both globally and in selected regions, by transferring concepts and adapting methods from digital stemmatic approaches.
The resulting genealogical connections are then interpreted using traditional historiographic methods, combining topographic and textual content with primary and secondary sources.
In this sense, we use computational tools to translate distortion patterns into interpretable forms and apply historiographic methods to interpret and narrate these findings. Through this process, akin to forensic analysis, the maps a
