From data to corpus: semiotic and documentary issues in audiovisual archives

TL;DR

This paper explores the theoretical, methodological, and technical challenges of constructing and analyzing audiovisual corpora in digital humanities, emphasizing semiotic frameworks, data documentation, and semantic enrichment.

Contribution

It provides a comprehensive analysis of the foundational issues in audiovisual corpus research, integrating semiotic theory with digital humanities methodologies.

Findings

Highlights the importance of a transdisciplinary semiotic approach.

Clarifies distinctions between data collections, corpora, and archives.

Discusses semantic enrichment processes for meaningful data analysis.

Abstract

The article examines the theoretical, methodological, and technical foundations of research on audiovisual corpora within the field of digital humanities. It outlines the main transversal issues underlying the processes of constructing, exploiting, and interpreting such corpora, which are conceived as specific forms of textual data in the broad sense - that is, as sets of semiotic traces (written, visual, sound, or multimodal) that make it possible to document, analyze, and transmit domains of knowledge. The analysis is organized around five complementary themes. The first concerns the status and structure of textual data lato sensu: any data, regardless of its medium, participates in a meaningful representation of a domain and therefore requires a unified theoretical and methodological framework based on a transdisciplinary semiotic approach. The second theme addresses the documentary value of data and corpora, understood as the relevance of materials for documenting a research object in relation to the goals and perspectives of the projects in which they are used. This value depends both on provenance and reasoned selection, and on the pragmatic context of their use. The third theme distinguishes between data collections, corpora, and archives. A data collection constitutes a potential reservoir of materials, whereas a corpus results from a reasoned and contextualized selection from this collection, elaborated in relation to a specific project. Archives, in turn, refer both to open data repositories and to research resources that support experimentation, editorialization, and valorization. This distinction highlights the dynamic nature of the processes of constituting and reusing digital resources. The fourth theme explores the semantic enrichment of data, understood as the set of semiotic and technical operations that give meaning to data, establish relationships among them, and make them usable by social actors. This approach raises issues related to the description, classification, and interconnection of data, notably through the use of ontologies, metadata, and models for knowledge representation. Semantic enrichment thus forms part of a broader reflection on value creation and data reuse in diverse contexts. Finally, the fifth theme addresses the instrumentation of research, that is, the digital environments, tools, and infrastructures that support the production, management, and dissemination of corpora. These technical systems - collaborative platforms, annotation, analysis, and publication tools - form an essential component of the digital humanities ecosystem and condition new forms of archiving, circulation, and knowledge transmission. In conclusion, the article emphasizes three central questions: how to construct meaning from data; how to model strategies of appropriation, reuse, and re-publication; and how to organize new semiotic and cultural ecosystems of research. The overall challenge lies in understanding how audiovisual data can be transformed into genuine objects of knowledge and value within contemporary scientific, technical, and cultural frameworks.