Challenges and solutions for the digital edition and geocoding of an 18th-century encyclopedia: the Diccionario histórico-geográfico de las Indias Occidentales
Résumé
Abstract This article presents the methodological and technical challenges encountered in the creation of a digital edition of Antonio de Alcedo’s Diccionario histórico-geográfico de las Indias Occidentales (1786–1789), combined with the extraction and geocoding of its geographic content. Our work sought not only to recover the dictionary’s spatial information as structured data for a historical gazetteer, but also to preserve the text as a coherent document, thereby avoiding its reduction to a mere reservoir of “spare parts.” We describe the dual workflow designed to accommodate both editorial representation in TEI-XML and semantic annotation of spatial entities, including the training of named entity recognition models adapted to noisy OCR and historical Spanish. Particular emphasis is placed on the challenges of geocoding colonial toponyms, where the availability of specialized historical gazetteers like HGIS de las Indias decisively shaped results, while experiments for areas without such a resource revealed the limits of semi-automatic matching. The solutions proposed for encoding entries, annotations, and normalized place data in TEI, though at times constrained by the standard, demonstrate the adaptive potential of TEI for projects at the intersection of text edition and data modeling. The resulting edition and online interface make visible both Alcedo’s vision of the Spanish Empire and the distortions inherent in his sources, offering scholars new means of interrogating colonial geographies and the knowledge practices of the late Enlightenment.