This special issue examines how computational and digital humanities methodologies can be mobilized from and with the Global South to preserve, interpret, and contest cultural heritage under conditions of uneven and heterogenous infrastructures, colonial archival violence, and epistemic inequality. Rather than treating “cultural analytics” as a narrow domain of model-driven pattern detection applied to pre-existing corpora, the contributions gathered here insist that, in many Global South contexts, knowledge production processes are directly dependent on the building of what is often treated as “data.” Such datasets are in practice actively constructed, situated, and selectively captured through socio-technical processes, and the building of collections, ontologies, and platforms is itself a primary site of analytical, theoretical, and political work. This framing foregrounds that what counts as data is not given in advance, but is shaped through the infrastructures, practices, and epistemic commitments that make its capture and interpretation possible. In this sense, what becomes legible as data is also conditioned by the designated communities imagined by a system, and by whether their knowledge practices, languages, and forms of representation are recognised within it (Pereda et al.).
Our point of departure is simple but consequential. The dominant templates for computational work in the humanities such as large, already-digitized corpora; mature NLP pipelines for hegemonic languages; stable repository infrastructures; and clear distinctions between “data production” and “data analysis” are historically specific to the Global North. When these templates are universalized as neutral standards of rigor, much of the most urgent work in Global South digital humanities is misrecognised as preparatory or just technical, in the sense of being treated as technical groundwork rather than analytical work in their own right. The essays in this issue collectively argue the opposite. Under conditions of archival fragmentation, linguistic marginalization, and heterogeneous infrastructural ecologies, including social, cultural and community-based infrastructures, the conceptual design of collections, the negotiation of community governance, the creation of reusable data and metadata, and the iterative assembly of corpora are central forms of analysis. They make particular worlds legible, contest others, and materially redistribute epistemic authority, often for the first time. In many cases, the issue is not simply the absence of infrastructure, but the structural misalignment of dominant tools and standards with the visual, relational, linguistic, and ontological logics through which knowledge is organised in these contexts.
Where Does “Analytics” Begin?
This tension becomes visible in the ways such work is sometimes misrecognised as digital collection projects rather than analytical studies, reflecting a narrower definition of analysis grounded in Global North epistemologies that separate data production from interpretation and invisibilise the analytical work embedded in infrastructural, technical, and community practices. In many cases, the tools, models and infrastructures developed within these paradigms prove only partially applicable—or require substantial adaptation, when working across different linguistic, cultural, and ontological worldviews. As a result, much of the work presented here does not simply apply existing computational methods but makes possible the emergence and articulation of forms of knowledge that have historically remained invisible or inaccessible within dominant technical frameworks, while aligning technologies and content with locally grounded practices and needs. This raises the question at the heart of this issue: at what point, and for whom, does analysis begin? In a well-resourced Global North archive with standardized metadata and abundant digitized sources, “analytics” can appear to start only once a clean dataset is in hand, and a model is run. In many under-resourced settings, particularly across the Global South, almost every step prior to modelling—including funding constraints, locating dispersed materials, negotiating access, digitizing fragile documents, stabilizing scripts and orthographies, co-designing ontologies with communities, and building interfaces that can function offline—is a site of epistemic decision-making and conceptual innovation. Consequently, cultural analytics, in these settings, emerges not from the application of data modelling alone but as a politically and epistemically situated practice, where the design of tools, workflows, and platforms actively shapes whose knowledge is represented, how it is structured, and under what conditions it resists being rendered as extractable data. To treat those activities as merely technical or preliminary is not a neutral methodological judgment; it reproduces the same hierarchies this collection seeks to unsettle. It assumes that analytical value resides only where familiar quantitative forms (graphs, topic models, embeddings) already appear, and it obscures the fact that for Indigenous and invisibilised communities, the most consequential analytic act may be the ability to name, connect, and preserve their own histories on their own terms. The contributions by Flores-Muñoz et al., Lakra et al., Vargas Montes and Rojas Solís, and C. Nguyen demonstrate that corpus-building, platform design, and interpretive modelling are inseparable in Global South computational humanities, and each paper theorizes and operationalizes this entanglement in a different medium and context.
The State of the Art: Infrastructures and Afterlives in the Global South
The application of computational methods to cultural heritage in the Global South stands at a critical juncture. On one hand, the past decade has witnessed a proliferation of digital humanities initiatives aimed at digitizing and preserving cultural materials from underrepresented regions, and today’s biodiversity databases, Indigenous language corpora, oral history archives, colonial document collections, and community-based digital repositories exist across Latin America, Africa, Asia, and the Pacific (Fiormonte; Risam; Pawlicka-Deger; Dutta). Major research institutions have begun developing machine learning models for tasks such as optical character recognition (OCR), handwritten text recognition (HTR), named entity extraction, and entity linking, which promise to accelerate the computational analysis of vast archival collections, including early modern and colonial records (Murrieta-Flores et al.; Candela et al.). International initiatives have also mobilized to standardize metadata schemas, develop interoperable linked-open-data platforms, and advocate for open-access publishing of cultural data (Candela et al.; Pawlicka-Deger).
Yet this growth masks persistent structural inequalities. The digital infrastructures, trained models, and technical standards that predominate in the field remain overwhelmingly products of research conducted in and for the Global North (Fiormonte; Risam). OCR and HTR models trained on European handwriting styles or well-resourced scripts struggle with non-Latin scripts and paleographies, while language technologies built for a small subset of economically dominant languages systematically under-serve the vast majority of the world’s 6,500 languages (Bird; Blasi, Anastasopoulos, and Neubig). Metadata standards and data documentation practices developed within Western archival and library traditions often prove ill-equipped to capture the relational, cosmological, and place-based ontologies that underpin many non-Western knowledge systems (Bender and Friedman; Candela et al.; Christen). This underscores that what circulates as “data” within digital heritage systems is shaped by specific technical standards, institutional priorities, and epistemic frameworks, which define not only how information is structured, but what forms of knowledge can be represented at all. Moreover, a substantial portion of digital heritage projects in the Global South continue to operate within top-down, extractive models in which external experts and well-funded institutions digitize materials belonging to marginalized communities and then publish the resulting “findings” in Northern venues, thereby reproducing the colonial division between knowledge producers and knowledge subjects (Fiormonte; Risam; Dutta). At the same time, however, these conditions give rise to alternative infrastructural configurations, often grounded in social, cultural, and community practices, which play a central role in sustaining and circulating knowledge beyond formal technical systems.
Technical infrastructure remains a defining challenge. While major universities and national institutions in capital cities may possess the technical capacity and financial resources to implement sophisticated digital repositories, smaller municipalities, rural communities, and grassroots organizations often lack reliable electricity, consistent internet connectivity, and technical support (Dutta; Pawlicka-Deger). This digital disparity ensures that preservation efforts remain concentrated in already-privileged centres, leaving many collections physically isolated and intellectually inaccessible, not because they lack epistemic value, but because the infrastructure to sustain their preservation simply does not exist or is not equitably distributed (Fiormonte; Pawlicka-Deger).
Linguistic diversity compounds these issues. The Global South encompasses the majority of the world’s linguistic diversity, yet computational resources such as taggers, parsers, machine translation, and language models exist for only a fraction of these languages (Bird; Blasi, Anastasopoulos, and Neubig). Many heritage materials are recorded in Indigenous, creole, or multilingual idioms that existing systems have never “seen,” and whose phonologies, scripts, and discursive forms resist straightforward integration into current NLP pipelines (Bird). Precisely those traditions most at risk of disappearance, such as Nahuatl oral narratives in Veracruz or Coahuiltecan linguistic traces in eighteenth-century Texas–Coahuila mission records, are often those least supported by computational tooling (Dutta; Murrieta-Flores et al.).
The challenge extends beyond technical capacity to questions of epistemology and power. Many digital platforms implicitly assume that heritage is best understood as discrete, catalogable units such as texts, images, and artifacts that can be decontextualised, standardised, and organised according to universal classification schemes (Fiormonte; Bender and Friedman). Yet in numerous Global South contexts, knowledge exists relationally; stories are inseparable from the landscapes they invoke, ceremonies are tied to seasonal and celestial cycles, intellectual property is communal rather than individual, and meaning emerges through oral performance and intergenerational dialogue rather than through fixed documentary forms (Dutta; Christen; Carroll et al.). Imposing Western archival logics on such knowledge systems risks doing epistemic violence by reifying what should remain fluid, atomising what should remain relational, and appropriating what should remain governed by the communities themselves (Risam; Fiormonte; Christen, Merrill, and Wynne).
Governance and ownership of digital heritage are equally contested. International instruments such as the United Nations Declaration on the Rights of Indigenous Peoples and related frameworks have helped articulate principles of Indigenous data sovereignty. Still, these rarely translate directly into platform architectures, licensing regimes, or institutional practices (Carroll et al.), often in contradiction with long-established indigenous rights and national legal frameworks. Many digitization projects still proceed without sustained community consultation; data is gathered under the rubric of “research” and then published in open repositories accessible to commercial and governmental actors who may extract further value from it (Christen; Risam). Even well-intentioned projects can reproduce extractive dynamics by centralising control of digital materials in distant institutions and limiting communities’ capacity to annotate, revise, or selectively share their heritage, a problem that Indigenous-designed platforms such as Mukurtu explicitly seek to address through community protocols and differential access controls (Christen, Merrill, and Wynne; Christen).
Against this landscape, the four contributions in this special issue offer models for an alternative computational humanities practice grounded in epistemic justice, data sovereignty, and community-driven methodology. Flores-Muñoz et al.’s Digital Itinerant Community-Based Library/Museum treats Nahua oral narratives as living, spatialized practices and develops a relational geospatial archive whose ontologies are co-designed with community members and Intercultural University partners, including strategies for offline access and local stewardship. Lakra et al.’s mapping of Purulia Chhau performances in rural West Bengal demonstrates how participatory, document-based GIS workflows can be built from diaries, receipts, and interviews, shifting spatial analysis away from institutional data silos toward collaborative reconstruction of itinerant performance circuits. Vargas Montes and Rojas Solís’ Indigenous Epistemes of the Borderlands constructs a digital collection that links colonial documents to Coahuiltecan-speaking nations through spatiotemporal and linguistic metadata, explicitly framing the database and mobility maps as tools for Indigenous ethnohistory, community use, and future linguistic revitalization. Nguyen’s interpretive work with Technique du peuple annamite shows how embeddings and clustering can underwrite a mode of cultural analytics that foregrounds multivocal authorship, structural absences, and speculative reading rather than positivist closure.
These projects also help us reframe the question of sustainability and the “afterlives” of digital work. Much Northern digital humanities has been shaped by a grant-driven project culture in which large sums produce highly polished, often centralised platforms that struggle to persist once funding ends, becoming what practitioners describe as data graveyards (VandeCreek). By contrast, many Global South initiatives operate with limited financial resources but are enmeshed in dense social and institutional infrastructures, including community organizations, intercultural universities, and public cultural institutions, that allow projects to endure, adapt, and circulate even in the absence of continuous funding (Dutta; Pawlicka-Deger). The Mexican national platform Mexicana, for instance, aggregates collections from museums, libraries, archives, broadcasters, and research bodies under the Secretaría de Cultura, offering long-term public access to radio programs, films, books, photographs, theses, and other cultural materials via a unified portal (Secretaría de Cultura; Sovarzo). The four projects in this issue resonate with such models of endurance. In doing so, they demonstrate that infrastructures are not only technical systems, but assemblages of social relations, institutional practices, and epistemic commitments that enable knowledge to persist and evolve.
Articles in the Special Issue
Modelling Oral Narrative in Central Veracruz, Mexico
“Modeling Oral Narrative in Central Veracruz, Mexico: A Digital Itinerant Community-Based Library/Museum” advances a digital infrastructure for Nahuatl oral narratives that treats corpus construction as analytical and decolonial work in its own right. Flores-Muñoz et al. draw on interviews with Nahua elders in the Sierra de Zongolica to prototype a “Digital Itinerant Community-Based Library/Museum” that integrates audio–video recordings, Nahuatl transcriptions, interactive mapping, and semantic cartograms to represent narratives as living territorial practices rather than static “folklore.” Methodologically, the project foregrounds the construction of a relational geospatial archive through community annotation workshops, co-designed micro-ontologies for Nahuatl concepts, and metadata that can encode ceremonial paths, affective landmarks, and non-linear temporalities. The contribution lies less in any single interface than in showing how digital collections in Indigenous languages can themselves function as arguments about what counts as data, how spatial humanities might accommodate non-Euclidean forms of place, and how computational tools can be made to listen to Indigenous epistemologies instead of subordinating them to inherited archival schemas.
Mapping the Intangible in Purulia Chhau
“Mapping the Intangible: Challenges and Innovations in the Spatial Analysis of Purulia Chhau” focuses on the itineraries of community-based Chhau dance troupes in West Bengal over a Bengali year, by using participatory, document-based mapping and GIS to reconstruct performance circuits from diaries, receipts, and oral testimony. Lakra et al. demonstrate that in data-scarce contexts, spatial analysis begins with the collaborative assembly of a corpus, geocoding place names, reconciling vernacular calendars with Gregorian time, and approximating village locations where official shapefiles do not exist. The article’s key contribution is to collapse the distinction between “collection” and “analysis”: reconstructing performance routes reveals how ritual calendars, agrarian cycles, transport infrastructures, and social media economies shape the geographies of Purulia Chhau. At the same time, the authors foreground the epistemic risks of flattening embodied, relational practices into discrete spatial points and propose “interpretive performance mapping” that treats ethnography, local spatial categories, and GIS as co-equal analytic modalities.
A Digital Collection for a Coahuiltecan Ethnohistory
“A Digital Collection for a Coahuiltecan Ethnohistory” addresses a classic problem in northern Mexican and Texas ethnohistory: the multiplicity of Indigenous group names and the difficulty of tracing continuities across fragmentary colonial archives. Vargas Montes and Rojas Solís’s project, Indigenous Epistemes of the Borderlands, builds a digital corpus that links eighteenth-century sacramental records and related documents to Coahuiltecan-speaking nations through systematic spatiotemporal and contextual annotation. Combining OCR, text mining, GIS, and network visualization, the authors identify particularly prominent nations in the Manual de Administración de Sacramentos corpus and develop a case study of the Alazapas’ mobility and political strategies. The database’s design, how it groups denominations, interprets movement, publishes bilingual digital editions, and interfaces with contemporary descendant communities, makes explicit that a “digital collection” can be an interpretive argument, an Indigenous-centred infrastructure for further research, and an intervention in debates on data sovereignty and borderlands historiography.
Interpretive Digital Humanities and Colonial Visual Archives
“Interpretive Digital Humanities: Social Worlds of Orality, Visual Communication, and Cultural Practice” offers a methodological intervention in Global South cultural analytics through a case study of Technique du peuple annamite (Kỹ thuật của người annam), a sprawling French colonial visual encyclopedia of Vietnamese crafts and everyday life. Working with thousands of drawings and multilingual captions, C. Nguyen employs visual and word embeddings, clustering, and minimum spanning tree visualizations not to extract definitive empirical claims, but to surface plural narrative threads, structural absences, and zones of unknowability within a violently produced colonial dataset. The article argues that corpus-building, including digitization, transliteration of Hán Nôm captions into quốc ngữ, and multilingual alignment, is already deeply interpretive. Subsequent modelling should serve question-generation, critical fabulation, and multivocal interpretation rather than “explainable” closure. In doing so, it reimagines cultural analytics as a speculative, relational, and dialogical practice attuned to the ethical stakes of working with colonial archives from the Global South.
Collections as Analytics, Analytics as Governance
Taken together, these four contributions demonstrate that the opposition between “digital collections” and “analytics” is conceptually misleading in Global South cultural heritage work. Each project makes visible how choices about what to digitize, how to name entities, how to encode space and time, how to structure access, and how to align technical workflows with community priorities are themselves forms of analytic reasoning and theoretical intervention.
In Veracruz, the itinerant library/museum’s architecture materializes a claim that Nahua oral narratives are spatial, relational, and ongoing acts of world-making rather than recoverable “texts” awaiting analysis. In Purulia, document-based participatory mapping shows that reconstructing the movement of Chhau troupes is already an analysis of how ritual calendars, rural economies, and media infrastructures co-produce a performance ecology. In the Texas–Coahuila–Nuevo León borderlands, relational databases and mobility maps contest long-standing historiographical assumptions by reframing Indigenous movement as an expression of cosmovision and sovereignty rather than as aimless wandering. In colonial Vietnam, embeddings and clustering are reimagined as ways to read against the grain of a colonizing visual regime and to foreground the partial, contested nature of any “totalizing” representational project.
In all four cases, computational tools do not appear at the end of the process as neutral engines of pattern discovery; they are folded into decisions about governance, pedagogy, language revitalization, and community benefit from the outset. This is precisely where the stakes of cultural analytics are highest in the Global South: when tools, workflows, and platforms are explicitly designed to redistribute epistemic authority and to resist the reduction of Indigenous and minoritized knowledges to extractable data.
Rethinking Cultural Analytics from the Global South
The special issue also foregrounds the importance of social and institutional infrastructures that sustain digital heritage work beyond project cycles. In Mexico, national initiatives like Mexicana demonstrate how publicly funded, centralized yet plurivocal platforms can bring together holdings from research institutes, museums, libraries, broadcasters, and archives, offering unified access to radio programs, films, books, photographs, and other cultural resources to diverse publics. Such platforms matter not only because they are technically robust, but also because they are embedded in state institutions, educational systems, and community practices that can sustain them over time, even in the absence of large, recurring grants.
The projects in this issue mirror this orientation toward durability. The itinerant library/museum model is conceived as replicable across schools and communities in the Sierra de Zongolica, with lightweight, offline-capable infrastructures that match local realities rather than idealized Northern standards. The Chhau mapping project is rooted in long-term ethnographic engagement and emphasises the need for governmental and institutional support to improve documentation practices among rural troupes, foregrounding how modest investments can transform the evidentiary basis for intangible heritage policy. The Coahuiltecan digital collection is explicitly designed as an evolving resource that can support current language revitalisation efforts and future ethnohistorical research, while acknowledging its limits and the need for ongoing community dialogue. Nguyen’s work, meanwhile, models how iterative, partially published, and “messy” digital humanities projects can invite broader publics and students into the interpretive process instead of waiting for a perfectly complete dataset or platform.
These examples suggest that Global South cultural analytics often derive their resilience from dense human networks including elders, students, archivists, community organisations, local universities rather than from large but finite streams of capital. To recognise this is not to idealise the lack of resources, rather, it challenges evaluative frameworks that privilege short-term technical sophistication over long-term social embeddedness, and invites journals, funders, and institutions to recalibrate what counts as a successful or exemplary digital humanities intervention. Recognising this requires a shift away from evaluating digital humanities work solely in terms of technical outputs, toward understanding how knowledge is produced through the interplay of infrastructures, communities, and epistemic practices.
By foregrounding projects that sit at the intersection of corpus formation, platform design, and computational interpretation, this special issue proposes an expanded understanding of cultural analytics that is accountable to Global South realities. In this view, cultural analytics is not a fixed set of tools applied to any cultural dataset, but a situated practice in which methods, infrastructures, and epistemologies are co-designed with communities, archives, and alongside languages and knowledges that have historically been marginalized.
For the Journal of Cultural Analytics, these contributions offer both a challenge and an invitation. They challenge any residual tendency to equate analytical value solely with large-scale, model-driven pattern recognition over already-stable datasets. They invite the field to recognize as central the work that precedes and surrounds such models. This includes the labour of building collections where none existed, of encoding indigenous and minoritized categories in metadata, of designing interfaces that can circulate within local infrastructures, and of theorizing computation as a medium for multivocal, speculative, and decolonial interpretation.
We offer this issue, then, as both a set of concrete case studies and the programmatic statement that in the Global South, computational approaches to cultural heritage necessarily blur the boundaries between collection and analysis, method and theory, infrastructure and critique. Attending that blur is not a compromise in rigor; it is a precondition for any genuinely global, just, and intellectually generative practice of cultural analytics.
Competing Interests
The authors have no competing interests to declare.
- Bender, Emily M., and Batya Friedman. “Data Statements for Natural Language Processing: Toward Mitigating System Bias and Enabling Better Science.” Transactions of the Association for Computational Linguistics, vol. 6, 2018, pp. 587–604, http://doi.org/10.1162/tacl_a_00041.
- Bird, Steven. “Bootstrapping the Language Archive: New Prospects for Natural Language Processing in Preserving Linguistic Heritage”. Linguistic Issues in Language Technology, vol. 6, Oct., http://doi.org/10.33011/lilt.v6i.1243.
- Blasi, Damián, Antonios Anastasopoulos, and Graham Neubig. “Systematic Inequalities in Language Technology Performance across the World’s Languages.” Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics, Association for Computational Linguistics, 2022, pp. 5486–5505, http://doi.org/10.18653/v1/2022.acl-long.376.
- Candela, Gustavo, Javier Pereda, Dolores Sáez, Pilar Escobar, Alexander Sánchez, Andrés Villa Torres, Albert A. Palacios, Kelly McDonough, and Patricia Murrieta-Flores. “An Ontological Approach for Unlocking the Colonial Archive.” ACM Journal on Computing and Cultural Heritage, vol. 16, no. 4, 2023, pp. 1–18, http://doi.org/10.1145/3594727.
- Carroll, Stephanie Russo, Desi Rodriguez-Lonebear, and Andrew Martinez. “Indigenous Data Governance: Strategies from United States Native Nations.” Data Science Journal, vol. 18, 2019.
- Christen, Kimberly. “Opening Archives: Respectful Repatriation.” The American Archivist, vol. 74, no. 1, 2011, pp. 185–210. JSTOR, http://www.jstor.org/stable/23079006. Accessed 20 Mar. 2026.
- Christen, Kimberly, Alex Merrill, and Michael Wynne. “A Community of Relations: Mukurtu Hubs and Spokes.” D-Lib Magazine, vol. 23, nos. 5–6, 2017, http://doi.org/10.1045/may2017-christen.
- Dutta, Uttaran. “Digital Preservation of Indigenous Culture and Narratives from the Global South: In Search of an Approach.” Humanities, vol. 8, no. 2, 2019, p. 68, http://doi.org/10.3390/h8020068.
- Fiormonte, Domenico. “Digital Humanities and the Geopolitics of Knowledge.” Digital Studies/Le champ numérique, vol. 7, no. 1, 2017, p. 5, http://doi.org/10.16995/dscn.274.
- Murrieta-Flores, Patricia, Rodrigo Vega-Sánchez, Alexander Sánchez Díaz, and Francisco Cruz-Ríos. “Unlocking Colonial Records with Artificial Intelligence: Achieving the Automated Transcription of Large-Scale 16th- and 17th-Century Latin American Historical Collections.” Science & Technology of Archaeological Research, vol. 11, no. 1, 2025, http://doi.org/10.1080/20548923.2025.2484828.
- Pawlicka-Deger, Urszula. “Infrastructuring Digital Humanities: On Relational Infrastructure and Global Reconfiguration of the Field.” Digital Scholarship in the Humanities, vol. 37, no. 2, 2022, pp. 534–550, http://doi.org/10.1093/llc/fqab086.
- Pereda, Javier, Pip Willcox, Gustavo Candela, Alexander Sanchez, and Patricia A. Murrieta-Flores. “Online Cultural Heritage as Social Machine: A Socio-Technical Approach to Digital Infrastructure and Ecosystems.” International Journal of Digital Humanities, 2025, http://doi.org/10.1007/s42803-025-00097-6
- Risam, Roopika. New Digital Worlds: Postcolonial Digital Humanities in Theory, Praxis, and Pedagogy. Northwestern UP, 2018.
- Secretaría de Cultura. “MEXICANA: Repositorio del Patrimonio Cultural de México.” Mexicana. Repositorio del Patrimonio Cultural de México, accessed 7 May, https://mexicana.cultura.gob.mx.
- Sovarzo, José. “Mexicana. Repositorio del Patrimonio cultural de México.” Blog de la Biblioteca Digital del Patrimonio Iberoamericano (BDPI), 8 May 2020, https://bdcv.hypotheses.org/3304.
- VandeCreek, Mark. “Great Lengths: A Review of Website Preservation Activities at Three American Universities with Digital Humanities Centers.” Digital Scholarship in the Humanities, vol. 39, no. 1, 2024, pp. 409–427, http://doi.org/10.1093/llc/fqad080.