Loading...
Loading...
A new paradigm for public library discovery systems—where ideas become first-class navigational objects, not just metadata attached to documents.
A visual introduction to idea-native discovery for public libraries
Public libraries help people find books. But what if they could help people find ideas?
Today's library catalogues are container-centric: they organise and surface documents (books, articles, records). Users must navigate to the right container to access the ideas within. The discovery question is: "Which book has what I need?"
Idea-Native Discovery inverts this. Instead of searching for containers that might hold relevant ideas, users navigate a landscape of ideas themselves. Concepts like "the sublime", "climate justice", or "colonial legacies" become navigable objects with their own identity, relationships, and manifestations across the collection.
The discovery question becomes: "What ideas am I exploring, and where do they lead?"
Library catalogues have always organised containers: scrolls, codices, records, digital objects. Metadata systems (MARC, Dublin Core) describe containers—author, title, subject headings. Discovery interfaces help users find containers.
But users don't want containers. They want what's inside: ideas, arguments, perspectives, narratives. The container is just a delivery mechanism.
What if the ideas themselves were navigable? What if "alienation in industrial society" was an object you could explore—seeing which works embody it, what ideas it connects to, how it evolved over time?
This doesn't replace traditional cataloguing—it adds a semantic layer above it. Containers remain, but ideas become the primary navigation surface.
| Aspect | Container-Centric | Idea-Native |
|---|---|---|
| Primary Unit | Document (book, article, record) | Idea (concept, theme, symbol) |
| Navigation | Find the right container | Explore the idea landscape |
| Relationships | Subject headings, keywords | Typed semantic relationships |
| Discovery Model | Search → Filter → Select | Enter → Explore → Connect |
| What Users Find | Items that match query | Ideas and their manifestations |
Conceptual comparison between container-centric discovery, where documents are the primary units of access, and idea-native discovery, where concepts and their relationships function as first-class navigational objects linked to source materials.
Key Transformation: Users navigate between ideas rather than documents, with source materials accessible as provenance links rather than primary discovery targets.
Schematic illustration of a local neighbourhood interface, showing a focal idea node, a bounded set of semantically typed related ideas, and user-controlled expansion paths.
Relationship Types
+24 more related ideas
The currently selected idea, displayed prominently with full context and source attribution.
Semantically labeled connections (supports, extends, challenges) enabling structured exploration.
User-controlled depth limiting prevents cognitive overload while maintaining discoverability.
Design Principle: The local neighbourhood interface respects human cognitive limits while preserving the graph's richness through progressive disclosure and semantic type filtering.
The paper proposes seven principles for designing idea-native discovery interfaces:
Ideas are disaggregated to the smallest pragmatically useful level—reusable across contexts.
Connections are named and navigable: 'embodies', 'contrasts with', 'evolved from'.
Every idea links to its source materials—the containers that embody it.
Users see a bounded graph of related ideas, not an overwhelming global network.
Complexity unfolds on demand—start simple, add depth as needed.
Relationships are visible and interpretable—users understand why ideas connect.
Enter via theme, symbol, era, or author—not just keyword search.
These charts illustrate the measurable differences between container-centric and idea-native discovery approaches.
How each approach performs across key discovery dimensions
Idea-native discovery dramatically outperforms on relationship visibility and serendipity
Importance vs implementation difficulty for each principle
Typed Relations and Semantic Legibility are highest priority; Multiple Entry Points is quickest win
Estimated discovery experience improvement for different library users
Curious browsers benefit most; subject experts see smaller gains (they already know what they want)
Three user journeys showing the difference between container-centric and idea-native approaches:
A postgraduate researching 'the sublime' in Romantic poetry
Container-Centric
Searches 'sublime romanticism', gets 847 results sorted by relevance, manually connects concepts across sources
Idea-Native
Enters 'the sublime' as an idea, sees its connections to 'terror', 'infinity', 'nature worship', explores how Burke's treatise influenced Shelley's poetry through typed relationships
A reader who loved 'The Great Gatsby' and wants to explore its themes
Container-Centric
Reads 'readers also borrowed' suggestions, or browses 'American Literature' section
Idea-Native
Enters via 'the American Dream', explores connections to 'class mobility', 'disillusionment', 'Jazz Age'—discovers Fitzgerald's other works, but also Steinbeck, Wharton, and contemporary novels that embody the same ideas
A high school student researching climate change for an essay
Container-Centric
Keyword search returns scientific papers, policy documents, news articles—overwhelming, no sense of how concepts relate
Idea-Native
Enters via 'climate change', sees idea map: 'greenhouse effect' (scientific mechanism), 'climate justice' (ethical dimension), 'adaptation vs mitigation' (policy approaches). Each idea links to age-appropriate sources.
This paper applies Idea-Native Architecture (INA) to a specific domain: public library discovery. INA is the broader framework for treating ideas as first-class governable objects. This paper shows what that means in practice for libraries.
Where INA provides the theoretical foundation (container-object separation, semantic agency, governance constraints), this paper provides concrete design principles for building discovery interfaces that embody those ideas.
Read the INA explainer →This paper is a working paper. It applies idea-native architecture to public library discovery systems.
Status: Working Paper
Subject headings describe what containers are about—they're container metadata. Idea-native discovery makes ideas navigable objects in their own right, with relationships, provenance, and their own identity independent of any single container.
Thesauri define term relationships for indexing consistency. Idea-native discovery uses semantic relationships that are richer (typed, directional, contextual) and designed for navigation rather than cataloguing. They're complementary but distinct.
The paper proposes a hybrid approach: expert curation for core concepts, community contribution for emerging ideas, and AI-assisted extraction for scaling. The key is starting with high-value ideas and growing incrementally.
Public libraries serve diverse users with varying information needs and literacies. They're also mission-driven institutions focused on access and discovery, not just retrieval. This makes them ideal contexts for rethinking how discovery works.
Explore the broader theoretical foundation in Idea-Native Architecture.
Idea-Native Architecture