Flow States: Designing Visually Fluid Visualisation Tools for Humanistic Enquiry

Humanistic datasets increasingly contain new levels of nuance and complexity, capturing the multi-faceted nature of human experience and the systems in which it is embedded. These complex dataspaces allow the possibility of new forms of exploratory investigation that traverse and blend diverse semantic contexts—spatial, temporal, and conceptual— that operate across zoomable scales. The traces made by these exploratory traversals may further be recorded to create sharable arguments and narratives.
Since data of this kind are often best modelled as semantically referenced knowledge graphs, whose nodes are linked at a density that risks appearing inextricably tangled, there is a pressing need for novel design approaches to interface development that render informational richness legible without oversimplification: for interfaces that are tractable to direct and multifarious manipulation but which also preserve complexity; resist reduction; support multiple viewpoints and encourage discovery and knowledge-making.
With these imperatives in mind, we propose Flow States, a suite of interactive visualisation prototypes designed to address this challenge: prototypes which enable exploratory investigation and render the informational richness of complex humanistic data legible to the user by means of lenticular and graphical distillation.
Worked through a range of practical examples that address the needs of diverse historical datasets—including census records, oral history recordings, business prosopographies and social history archives—these prototypes propose novel interface design approaches which facilitate exploration, discovery and knowledge-making. In this presentation we will: demonstrate practical examples and case studies; outline the design principles and concepts underpinning the development of these tools; and discuss their general applicability across a variety of human-centred data contexts: e.g. social networks, life events, written transcriptions, and annotated audiovisual materials
The current core of Flow States compromises three components: the Alluvial Visual Filter, the Annotated Time-Based Media Negotiator, and the Temporal-Social Relation Explorer.
• The Alluvial Visual Filter is a visual query builder that enables intuitive upstream exploration and filtering of large and unwieldy datasets, via ‘sluice gates’: these allow users to refine the flow of data attributes, iteratively and provisionally, and reveal groups of objects or individuals that share similar characteristics, in order to surface patterns of identity, role, or aLiliation within a dataset.
• The Annotated Time-Based Media Negotiator is a graphical media viewer designed for granular, zoomable exploration and interpretation of pre-annotated audiovisual materials via an interface which supports comparison across multiple media sources.
• The Temporal-Social Relation Explorer enables users to explore social relations in a number of diLerent circumstances, including evolving household structures over time (using census data) and the nature of specialist businesses as they developed from the artisanal to industrial practice (referencing personal and institutional records, and detailed properties of these) - revealing both individual life trajectories (age, job, status,) and broader patterns—composition, leadership, familial and professional connections.
Each of these components are conceived and built as independent, yet interoperable data visualisation elements which can be used in combination, or in conjunction with others, to assemble coherent exploratory visualisation environments according to the potential and needs of particular datasets. At present the Alluvial Visual Filter has been integrated with the Media Negotiator and Temporal-Social Explorer as a pre-configurational tool. However, a fully combined environment that brings together both social-structural and audiovisual materials holds the potential for deepening interpretive insight. This is one of several next stages of work, whose active development will be discussed.
With a focus on case-studies that involve these components, this presentation will consider design principles that have a generalisable applicable to a range of data which represents human-centred scenarios. It will outline a set of core visualisation design considerations that are fundamental to this approach: i.e. ‘single view -multiple state’ interfaces, animated transitions to support a ‘subject in transit’ perspective; visually flexible and zoomable views; a coherent sense of proximity to the data source; interaction- richdataobjects;andtheprecedenceof‘making’over‘showing’.
Flow States is not a completed set of work, but a series of evolving investigations into what humanistic data interfaces may look like. We aim to situate this work as both as a design provocation and a set of practical design contributions. The presentation calls for a reconsideration of where and how visualisation is positioned within humanistic research practice. This session will, therefore, stand as a work-in-progress reflection, inviting dialogue around its aLordances, challenges, and future trajectories.
Attendees of this presentation will gain: a framework for thinking about exploration and discovery in humanistic data visualisation; insight into novel approaches for visualising human-centred data in practice; strategies for designing interfaces that support narrative multiplicity and legibility; design principles for creating exploratory, interoperable research environments.
This work is particularly resonant with the Information+ community because it sits at the intersection of data, design, and human-centred inquiry. It contributes to ongoing conversations about how we represent non-quantitative knowledge, and how we engage complexity without simplification.