The semiotic turn in digital archives and libraries

Abstract

This article explores the theoretical and methodological foundations of a semiotic approach to the design, organization, and use of digital archives, particularly audiovisual ones. It situates this reflection within the broader evolution from an information society to a knowledge society, emphasizing that digital data become meaningful only through processes that transform them into cognitively relevant resources.Digital media objects—texts, videos, sounds, or images—are not inherently knowledge resources but potential ones. They must undergo qualitative transformations such as classification, segmentation, indexing, annotation, translation, and republication to become meaningful for specific user communities. This semiotic processing embodies the shift from data to metadata, from simple information to structured and usable knowledge. The key issue is thus the semiotic structure of digital data: how meaning is organized, mediated, and made functional within given contexts of use.Using the French Audio-visual Research Archives (ARA) as a case study, the author describes an online repository of more than 6000 hours of scientific video content in the humanities and social sciences. Despite its richness, the ARA faces three central challenges:1) Linguistic limitations — most content is monolingual and therefore inaccessible to non-francophone audiences.2) Hidden information — meaning remains implicit without systematic description and indexing.3) Contextual adaptation — materials must be reformatted for specific user needs, especially in educational and intercultural contexts.To address these challenges, ARA engages in digital repurposing: reauthoring and adapting audiovisual materials through processes such as re-segmentation, re-description, and re-publication using diverse genres (e.g., hypermedia video books, thematic portals, bilingual folders). Figures 2–5 in the article illustrate these repurposing practices, showing how raw video data are transformed into pedagogically and culturally relevant digital artifacts.The paper introduces the Studio ASA (Atelier de Sémiotique Audiovisuelle), a digital environment comprising four specialized “workshops” for processing, indexing, publishing, and modeling archives. It operates through a meta-language integrating ontologies, thesauri, and description models that systematize how meaning is represented. This framework supports what the author calls a text- or discourse-based vision of digital archives, where each media object is treated as a structured text composed of layers (topical, rhetorical, expressive, and functional).Drawing on traditions from structural, functional, and cultural semiotics (Greimas, Halliday, Lotman, Peirce), the author defines semiotics as a methodology for understanding how meaning is produced, exchanged, and conserved through media. Figure 9 (p. 80) outlines four domains of semiotic application:1) Archive project definition (scope, discourse universe, meta-language design);2) Archive production (corpus creation, recording, long-term preservation);3) Re-processing and indexing (description, translation, interlinking);4) (Re-)publishing (authoring and distributing digital objects through specialized genres).The “semiotic turn” proposed here redefines digital archiving as an interpretive and authorial process rather than a merely technical one. It advocates integrating semiotic analysis with knowledge engineering to bridge data structures and human interpretation. In this view, digital archives become dynamic cultural ecosystems where data are continuously recontextualized through linguistic, cognitive, and social mediation. The article ultimately positions semiotics as both a methodology for digital knowledge production and a conceptual framework for intercultural communication in the global information environment.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 13 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.