Premodern urban economy relied massively on annuities as a (mostly) real-estate-based credit instrument, turning houses into an important economic factor. Sources on transactions related to real estate are generally dispersed in the archives – and any finding is hard to localize precisely, which is probably why real estate has played a minor role in research on the premodern urban economy. Finding means to research such transactions on a large scale allows us to discover a dynamic field of economic activity.

The ability to address such questions for Basel is based on the Historical Land Register of Basel, known as “Historisches Grundbuch Basel”, initiated in 1895 and developed over several decades (see https://dls.staatsarchiv.bs.ch/records/1016781). The creation of this land register involved a thorough plowing through large parts of Basel’s city archives. Each mention of a house prompted the creation of a file card, which contained an almost verbatim transcription of the source, accompanied by a date and additional details. These cards, approximately 120,000 in total, were organized by house and sequenced chronologically. Due to this arrangement, any discovery within the register can potentially be localized and dated. The wealth of information within the historical land register is unparalleled for the era since it is not limited to one corpus but contains combined information from various corpora.

For the Spatial Humanities 2024 conference, we want to discuss two significant challenges: one related to the process of georeferencing data and one to strategies for data analysis in space and time.

Localizing houses

When the Historical Land Register was established, it was decided to base its structure on an 1862 address book of the city of Basel. That was a pragmatic decision that allowed for an easily manageable structure for researchers. However, this structure was faulty for obvious reasons: it could not consider the changes that occurred during several centuries of building activity and real estate trading. When sources indicated that plots were divided or united, the editors of the Historical Land Register created new house dossiers linked to the exact address of 1862. This has the advantage that the dynamics of plots and housing can be retraced at the cost of having multiple dossiers linked to one address. These dossiers were then described as being “part of” plots as they presented themselves in 1862 or having multiple addresses in 1862 in case of plots that were split later. Based on this metadata and the dates we can extract from the records, we can establish some kind of “plot history” for the whole city (see an example for one street in the annexed document). By the way, this plot history is one of ownership rather than one of construction. For example, a person might have bought a neighboring house at some point and built a new house in place of the two old ones at some other time. This last activity generally cannot be identified in the sources.

The paper will present our current strategies for establishing a plot history. We intend to use rule-based procedures based on metadata to displace points in probable directions. The intention is not to find precise locations for houses that no longer exist but to create plausible approximations. The handling and representation of uncertainty is a major challenge in this process.

Spatial analysis

Having established a usable localization for the houses in Basel, we can start analyzing the real estate market. Based on procedures that will not be presented here, we intend to identify the primary transaction of each record in the Historical Land Register using machine learning methods. In a preliminary analysis, we tried to identify seizure procedures. Using a regex search in the HTR-recognized texts, we found about 9’500 such procedures from around 1350 to 1800 AD. In the annexed document, we show two representations of these data: plotting all points on a 19th-century city map and using a clustering algorithm by QGIS.

For the Spatial Humanities conference, we will explore various spatial analysis strategies for such data and challenges linked to representing the results in space and time. In addition to using analysis tools and procedures, we will focus on two aspects. Firstly, we explore the use of known physical and cultural structures of the city space (such as streets, suburbs, church parishes, etc.) to structure our data or their representation. Secondly, we intend to revert the process, using our data to determine the city's relevant structures and economic hotspots.
For this part of our paper, we will present the preliminary result of an exploratory phase in our research.

The Spatial Humanities Conference 2024 will be the ideal occasion to present such findings based on a rich data set to a competent public, allowing us to develop our research further!