Visualising uncertain and conflicting Cultural Heritage data

Epistemic uncertainty is a foundational condition of humanities data, often emerging across multiple interrelated dimensions with distinct evidential and hermeneutic implications. Despite the extensive literature on uncertainty visualisation in the Digital Humanities (DH), practical implementations remain scarce, and web-based DH projects rarely treat epistemic uncertainty as a primary organising principle. While the lack of adequate representation risks oversimplifying complex interpretative information and creating an illusion of certainty, representing multiple uncertainty dimensions increases visual and cognitive complexity. Drawing on the case study based on the data of the Federico Zeri Foundation photographic archive, which catalogues author attributions across a broad corpus of artworks, this contribution presents ongoing doctoral research addressing two questions: (RQ1) Can epistemic uncertainty function as the primary lens for enquiry and decision-making in cultural heritage visualisations? (RQ2) Can we use narration as a technique to support the interpretation and disambiguation of multiple epistemic uncertainties in an otherwise solely explorative and visually complex setting? The dataset offers a valuable context to investigate how epistemic uncertainty can be meaningfully represented and communicated in digital environments. Multiple, often conflicting attributions coexist with heterogeneous evidence and qualifiers expressing varying degrees of doubt, forming debates of different scales. To address this complexity, we adopt a bottom-up, iterative approach based on co-design sessions with domain experts, investigating how to integrate explorative visualisations with narrative layers that guide users through uncertainty and support interpretive engagement. In addressing the cognitive and visual challenges arising from the simultaneous representation of multiple uncertainty dimensions, narration is thus positioned as a mediator between data complexity and informed scholarly decision-making. This research contributes to ongoing discussions at the intersection of information visualisation and DH, questioning the representational limits of visual solutions for epistemic uncertainty and exploring its potential treatment as a first-class concern in the design of cultural heritage visualisation systems.