Public discourse shapes how societies understand complex topics like climate change, migration and identity. Yet while research has generated rich insights into framing, media effects and polarization, we still lack integrated ways of tracing how narratives take hold and compete across time and institutional settings (Brüggemann & Meyer 2023; Kreiss & McGregor 2024). This paper presents HONE (How Narratives Emerge), a theoretically motivated research project that deliberately engineers its methods and infrastructure to serve the public sphere. While its core scientific question is whether complex systems approaches can model language effectively, the project is designed from the outset around public-produced material, civic relevance and the development of engagement-enabling research infrastructures.

This paper argues that engaged digital humanities requires more than participatory formats alone (e.g. citizen science, co-creation initiatives). It also requires conceptual and technical infrastructures that make society-generated data legible in practice, i.e. engaged research must develop methods, models and interfaces capable of rendering complex social and cultural processes observable, interpretable and discussable. HONE is presented as a concrete case of how such infrastructures can be deliberately designed.

HONE takes society-produced meaning-making as its core object. It models public discourse as a complex system, but develops ad hoc approaches suited to public-generated language rather than importing models built for numerical or behavioral data. Narratives, framings and rhetorical choices are treated not as surface features, but as constitutive dynamics through which public discourse evolves.

This orientation is reflected in the project’s grounding in digital collections from European memory institutions, including parliamentary speeches, digitized newspapers and broadcast metadata. These archives are approached not simply as repositories of texts, but as records of collective meaning-making through which societies construct memory, negotiate identity and frame political reality.

Methodologically, HONE is built as an interdisciplinary infrastructure rather than a standalone analytical pipeline. Linguists, digital humanists and complexity scientists co-design how discourse is described and modeled. Linguistic and cultural expertise is embedded directly into the research infrastructure itself: in the definition of annotation categories, in decisions about what counts as a narrative unit, and in how discursive features are operationalized for modeling.

An integral component of the project is the development of public-facing visual models that make discursive dynamics observable. These include interactive network views of how framings move between speakers and institutions, and temporal maps showing how narratives rise, stabilize or fragment. Because complex systems dynamics such as emergence, non-linearity and amplification routinely defy intuitive understanding (with tangible societal consequences, such as misinterpretations of pandemic spread) HONE treats visualization not as dissemination, but as engagement infrastructure. These interfaces are designed to support public discussion, allowing users to trace how small rhetorical shifts can accumulate into large-scale changes.

HONE demonstrates how theoretically grounded digital humanities research can be deliberately designed with a civic function in mind: to support public sense-making and institutional mediation in a discursive environment increasingly shaped by polarization and misinformation (Kubin & von Sikorski 2021; Muñoz et al. 2024).