REWIND · Rereading European Cultural Heritage in Latin American Women Writers’ Travel Literature of the early 20th century: contrasting testimonies to build inclusive historical discourses

Rereading European Cultural Heritage in Latin American Women Writers’ Travel Literature of the early 20th century: contrasting testimonies to build inclusive historical discourses

The REWIND project is an international initiative that employs an innovative, interdisciplinary methodology rooted in Digital Humanities. It uses travel literature as a historical source to explore European cultural heritage through decolonial and gendered perspectives. The project's workflow integrates XML-TEI encoding, GLAM-Linked Open Data, and Geographic Information Systems to develop an ontological model connecting early 20th century European cultural heritage with the semantic web. Additionally, aspects of the reception of dissonant heritage were analysed using techniques such as sentiment analysis, collocation studies, and semantic field mapping, among others from Corpus Linguistics. The project's aim was to amplify the voices of Latin American women engaged in feminism and sociocultural diversity, positioning them as agents in constructing historical narratives about European cultural heritage. This approach confronted the authorised heritage discourse, promoting decolonial and inclusive historical narratives. By exploring diverse historical discourses, the project adopted a polyphonic approach to the reception of heritage. It went beyond simply adding women's testimonies or listing cultural productions they created, aiming to construct a more plural history. However, it faced challenges in countering the authorised heritage discourse of the period. Identifying the authors as Latin American women does not automatically exclude perspectives influenced by whiteness, androcentrism, patriarchy, or Eurocentrism, which may reflect their 'European' education. This complexity was further compounded by the elitism present in certain social circles, which constituted the primary Latin American demographic able to undertake European leisure travel during the first decades of the 1900s. The REWIND project structured into specific research and innovation objectives [R&IO]: R&IO1: To reconstruct the travellers' itineraries [WP2]. Geo-semantic annotation of travel books by identifying and mapping the European cultural heritage elements mentioned by authors. The XML markup language and the standards proposed by the Text Encoding Initiative (TEI) were used to annotate geo-semantically the books in the REWIND project corpus. To ensure interconnected information, each cultural heritage element was assigned a unique identifier using the @xml:id attribute, linked to an index with detailed information about each item through the @ref attribute. This information included geographical coordinates, typology, URI, and other relevant data. The encoding procedure was thoroughly documented in the file REWIND_Encoding_Guidelines.pdf and is available in the project's Zenodo repository. R&IO2: To connect the information using Linked Open Data (LOD) [WP2]. An ontology based on CIDOC-Conceptual Reference Model. The REWIND_Ontology.owx was designed using the Web Ontology Language (OWL), a markup language for publishing and sharing data through ontologies on the web. The OWL language aims to facilitate a markup model built on RDF and encoded in XML. The model CIDOC-CRM was employed to ensure data interoperability, providing an extensible ontology for concepts and information in cultural heritage and museum documentation, as travel books often reflected a traditional conception of heritage. Additionally, the controlled vocabulary from the ROSSIO Infrastructure was used to create typology entities corresponding to the E55 Type class, the GeoNames gazetteer for entities in the E53 Place class, the VIAF authority file for entities in the E21 Person class, and Wikidata for all identified cultural heritage elements mentioned in the studied travel books, associated with E22 Human-Made Object and E28 Conceptual Object classes. The XML encoding was based on the ontological model, so exporting the information from the books in a machine-readable format to the ontological model was straightforward. This approach enabled the study of cultural heritage items through patterns and relationships that are difficult for humans to detect. The mapping procedure was thoroughly documented in the file REWIND_Ontology_Guidelines.pdf and is accessible in the project's Zenodo repository. R&IO3: To recognise impressions associated with the European cultural heritage [WP3]. Sentiment analysis is applied to explore the evocative vocabulary of the cultural heritage descriptions. A sentiment analysis was conducted on all cultural heritage elements potentially associated with the concept of dissonant heritage, such as the Columbus monuments in Barcelona, Genoa, Seville, and Madrid, due to their connection to the colonisation of the Americas, as well as other elements related to political issues, such as the French Revolution. In this context, the results related to the Monastery of El Escorial and the statue of Isabella II of Spain were particularly interesting. However, it was identified that this type of approach presented biases. The NRC Emotion Lexicon (Mohammad & Turney, 2013) was specifically chosen because it had been manually annotated. Still, for example, when analysing the Mosque of Córdoba, the emotional tone was consistently negative, despite the text praising the historic building. This issue led to adjustments in the analysis, such as studying the semantic fields of words associated with the eight 'basic emotions' defined by Plutchik (2001) and the lexical placement of adjectives. Ultimately, it was decided not to share the scripts and workflow on Zenodo to prevent the identified biases from being perpetuated in future studies. R&IO4: To determine the relationship between European cultural heritage and otherness processes [WP3]. Gender decolonial theories emphasise the need to make non-European women visible as cultural agents in historical narratives to break the Eurocentric, patriarchal, and androcentric cultural canon. There is no single historical narrative; however, a predominant one often emerges depending on the political context. The REWIND project analysed the general historiographical trends documented in the most impactful publications on cultural heritage between 1900 and 1930, alongside information from popular travel guides by Baedeker (1898), Bradshaw (1894), and Murray (1890). Once the highlighted monuments or works of art were identified within authorised heritage discourse, along with the historical events or processes with which they were associated, it became possible to compare this data with references in the studied travel books to explore discrepancies between official and hegemonic historical narratives and the reception of Latin American women writers.

Data: CORDIS, © European Union