This short paper presents an ongoing project that explores the art criticism of Mexican feminist artist Mónica Mayer, based on 999 newspaper columns she published in El Universal between 1988 and 2008. Through a collaborative approach involving the artist herself, we analyze how Mayer constructed discourses around feminism, institutional critique, and the cultural politics of art in contemporary Mexico. Our work reactivates these writings as a digital feminist archive, using computational textual analysis to critically engage with questions of memory, authorship, and power.

The project is grounded in feminist theories of archival intervention (Eichhorn 2013), situated knowledge (Haraway 1988; Harding 2021), and critical feminist historiography (Pollock 1999; Nochlin 1971). These frameworks guide our understanding of Mayer's writing as a political and affective form of memory-making: both a critique of dominant art institutions and an act of feminist world-building through language. The archive is approached not as a static record, but as a living site of cultural struggle and reactivation.

In alignment with the conference theme Remembering, the project treats the corpus as a feminist cultural artifact whose digital remediation opens new interpretive possibilities. Methodologically, we engage with the tension between distant and close reading (Moretti 2005), combining topic modelling and co-occurrence analysis with network visualizations in Gephi. However, these tools are deployed within a critical framework that resists reductive pattern recognition. Drawing on Christof Schöch’s discussion of data in the humanities (2013), we are attentive to the distinctions between “big,” “smart,” and “messy” data, approaching our corpus as carefully curated “smart data” that reflects theoretical, methodological, and cultural commitments.

To build a traceable and critically assembled corpus, we gathered 999 columns authored by Mayer for her weekly Artes Visuales column in El Universal. The sources included printed versions, original faxes, digital files, and newspaper clippings, spanning from August 1988 to June 2008. Given this heterogeneity, corpus creation followed a structured workflow involving scanning, OCR, format conversion, and manual editing. Text was extracted using Adobe and Transkribus OCR systems, depending on the material’s quality, and was post-processed for readability while preserving editorial fidelity. Each column was represented as a plain text (TXT) file, paired with metadata (date, title, source type). This data was consolidated in a spreadsheet containing date, title, and full text, forming the analytical foundation. The decisions behind file selection, OCR correction, and canonical version designation were made collaboratively and documented to ensure transparency. These steps foreground the idea that the corpus is not a neutral dataset but a constructed, interpretive artifact shaped by technical, editorial, and ethical decisions.

Our analytical approach centers on an intensely collaborative form of distant and close reading, in which Mayer not only authors the corpus but also actively participates in its analysis. Her dual role as subject and co-analyst challenges conventional divisions between researcher and researched, human and machine, object and subject. This process becomes a form of embodied, situated interpretation where digital tools scaffold a critical, co-authored narrative. This methodological configuration constitutes a “very close reading,” rooted in dialogue, care, and critical reflexivity, and aligns with the feminist imperative to question knowledge hierarchies and center marginalized epistemologies.

While Named Entity Recognition (NER) has also been applied to the corpus, those findings are part of a separate publication. In this paper, we focus on the thematic and relational maps generated through topic modelling and network analysis to explore how feminist discourses take shape, circulate, and persist across time and genre. During the presentation, we will share visualizations created using Gephi to illustrate the networked relationships among actors, institutions, and discourses within the Mexican art scene. In addition, we will present results from topic modelling used to identify and interpret thematic communities that emerge across Mayer’s corpus, providing a multifaceted portrait of the cultural and political dynamics captured in her writings.

We conclude by reflecting on the epistemological and ethical stakes of conducting digital feminist scholarship with small, non-anglophone, artist-centered datasets. By emphasizing collaboration, contextual sensitivity, and archival responsibility, this project contributes to a broader understanding of how digital humanities can engage meaningfully and critically with feminist memory work, reimagining data not as an abstract representation but as a situated cultural production.