Problem
Most websites treat links as navigation. A link sends a visitor somewhere else: a menu item, a call to action, a source, a related page. That is useful, but it is a limited understanding of hypertext. Hypertext is more than movement between pages. It is a way of organizing thought through associations, references, trails, revisions and context.
The problem for a knowledge site is that links can become decorative. A page may link to many things without explaining why those things matter. A footer may list related content without revealing the conceptual relation. Search engines and AI systems may discover URLs, but not understand the site as a body of knowledge.
Electronic Artefacts needs a stronger model. The Knowledge Hub should use hypertext as an augmented knowledge system: a public interface where concepts, projects, programs, sources and articles reinforce one another through stable identity and typed relation.
Introduction
Hypertext begins with a simple proposition: knowledge does not have to be read only in a fixed sequence. A reader can move from one idea to another by association. A publication can become a network. A source can point backward to history and forward to new interpretation.
For Electronic Artefacts, hypertext is the public reading layer of the knowledge graph. The graph gives entities stable identities. Hypertext gives those entities a human path. A concept page can lead to a project. A project can lead to a technical article. A technical article can lead to sources, tools, methods and related concepts.
That structure is also a theory of augmentation. A good knowledge system does not only store information. It helps a reader perceive relations they would not otherwise notice.
Context
Vannevar Bush’s memex imagined associative trails as a response to the growing difficulty of managing scientific and cultural records. Douglas Engelbart’s augmentation program pushed further: computation could improve human intellectual effectiveness by changing how people represent, manipulate and share complex structures.
The web made linking ordinary, but ordinary links did not automatically solve the problem. Web pages are often chronologically buried, URL structures change, and link meaning is rarely explicit. A knowledge graph adds the missing layer: this page documents a concept, that project applies it, this article cites a source, that program implements a framework.
History
Hypertext is part of a longer history of memory systems, reference works, indexes, encyclopedias, scholarly citation, card files and annotation. Bush and Engelbart are important because they connected association with technical systems. They did not describe a mere library catalogue. They described work environments where links, representations and tools change the user’s capacity to think.
The web standardized a global linking environment. URIs made resources addressable. HTML made anchors ordinary. Search engines made pages discoverable. Later wikis, blogs, digital gardens and graph note systems made linking personal and editorial.
Electronic Artefacts stands in this lineage when it treats the site as a living research library rather than a promotional archive.
Core concepts
Node: a meaningful unit, such as a concept, article, project, source or program.
Link: a traversable connection between nodes.
Trail: a curated sequence of links that helps a reader follow an argument or topic.
Anchor: the visible or machine-readable point where a relation becomes reachable.
Context: the surrounding explanation that tells the reader why a link exists.
Identity: the stable address of the thing being linked.
Architecture
A durable hypertext system needs four layers. The first is URL design: pages must have stable, readable, canonical routes. The second is entity identity: the page and the thing described by the page are related but not identical. The third is relation semantics: links should be supported by graph predicates where the relation is important. The fourth is editorial curation: suggested readings and related articles should be selected for meaning, not merely generated from shared tags.
In Electronic Artefacts, this architecture already appears through canonical pages, identifier routes, JSON-LD, relation files and graph neighborhoods. The next step is editorial density: more articles must use these structures intentionally.
Implementation
Implementation begins in frontmatter. Every article should name subjects, sources, claims and evidence. The body should then explain those relations in language. A link to VASTE should not appear because VASTE exists; it should appear because the article is discussing runtime systems, contextual execution or graph-based infrastructure.
Internal links should point to stable routes. Concept mentions should become concept links when they help the reader. Articles should include “Related concepts” and “Related articles” sections, but those sections should be curated rather than generic.
At build time, the site can generate graph neighborhoods and search documents. At writing time, the editor supplies judgment.
Practical applications
Hypertext supports onboarding. A reader who arrives through search for “knowledge graph” can move into Linked Data, entity identity, VASTE and Runtime Theory.
It supports research. A reader studying signal archaeology can follow links into provenance, digital preservation, ORETH and Palimpsests.
It supports AI retrieval. Structured links, headings and JSON-LD make the site easier to parse without reducing articles to keyword pages.
It supports memory. Future Electronic Artefacts work can attach new projects to older concepts instead of creating isolated announcements.
Tools
Useful tools include static site generation, Markdown with structured frontmatter, JSON-LD, RDF concepts, sitemap generation, graph validation, search indexes, backlinks, source records and content audits.
Evidence
Electronic Artefacts already has the necessary foundation: entity records, typed relations, canonical routes, identifier routes and generated graph views. The new Knowledge Hub articles make that infrastructure visible to readers.
The strongest evidence is navigational: a reader can move from this article to Hypertext, Augmented Intelligence, Knowledge Graph and Linked Data without leaving the conceptual frame.
Editorial method
The editorial method is simple but demanding. Each page should answer three questions: what is this about, what does it connect to, and why does that connection matter?
Links should be added when they improve comprehension. A link can define, cite, contrast, extend or contextualize. If a link does none of those things, it is probably noise.
Common mistakes
The first mistake is using links as decoration. A page with many unexplained links may look rich while remaining conceptually shallow.
The second mistake is confusing tags with relations. Tags group pages by broad similarity. Relations describe specific claims.
The third mistake is letting time dominate structure. Chronological feeds are useful, but a research library should privilege concepts, sources and durable questions.
Electronic Artefacts implications
Electronic Artefacts should treat every major article as a node in a long-term knowledge system. VASTE, ORETH, Palimpsests and Vestiges should be linked when they illuminate a subject, not when promotion is needed.
The site’s advantage is not only its projects. It is the ability to explain the technical, artistic and historical ideas behind those projects in a public, reusable graph.
Future work
Future articles should cover wikis, digital gardens, annotation systems, URI design, backlinks, semantic publishing and the relationship between public knowledge pages and AI retrieval.
Related concepts
Read Hypertext, Augmented Intelligence, Knowledge Graph, Linked Data and Entity Identity.
Suggested reading
Start with Bush and Engelbart for historical grounding. Then read RDF, URI design and Linked Data material to understand how hypertext becomes semantic infrastructure.
Related articles
Continue with Linked Data and Public Knowledge Pages and Personal Knowledge Systems and Digital Gardens.
Glossary
Hypertext: linked text or media designed for non-linear traversal.
Trail: a meaningful sequence of connected nodes.
URI: a generic identifier for a resource.
Augmentation: expanding human capability through tools, representations and methods.
Limitations
Hypertext can fragment attention. Not every article should become a maze. A strong page needs a clear argument and carefully chosen exits.
Links also decay. Long-term hypertext systems need redirects, stable identifiers and periodic link review.
References
- Vannevar Bush. As We May Think. 1945.
- Douglas C. Engelbart. Augmenting Human Intellect. 1962.
- RFC 3986. Uniform Resource Identifier Generic Syntax. 2005.
- Electronic Artefacts. Knowledge Graph and Linked Data records.