For full text search please use the '?' prefix. e.g. ? Onboarding

Referent

Definition

The referent is the thing (real or imaginary) to which a resource’s identifier refers. Every identifier denotes its referent,

Node vs. referent

  • Referent: the subject that the node’s identifier names (a person, concept, event, dataset, etc.).
  • Node: the mesh construct that provides an identifier for and contains linked data about the referent. It also provides linked data about itself (as a node), and may contain other resources used for supporting Semantic Flow sites.

To talk about the node itself, you use its node handle (e.g. published IRI https://ns.example.org/persons/djradon/_node-handle or mesh identifier <djradon/_node-handle>).

Where it’s described

  • The referent’s description lives in the node’s reference flow.
  • The node’s own metadata and provenance live in the _node-* flows (e.g. _meta, _node-config-*).

Special case: payload nodes

  • In a payload node, the referent is not an external entity but an evolvable dataset contained in the node.

  • The dataset evolves as versioned distributions inside the node’s _payload (e.g. v1/, v2/, …).

  • The _ref may describe the dataset, e.g. its name, type, and provenance.

  • Example:

    • Node IRI: https://ns.example.org/projects/atlas/
    • Referent: the Atlas dataset (identified by the node IRI, evolving over time).
    • _ref: declares it as a dataset, supplies label and attribution.
    • _payload: provides concrete versions (v1, v2, …).

Why referent matters

Understanding what a IRI refers to is crucial for proper semantic web implementation. In the past, people have tried to use content IRIs to represent the things they refer to. A classic example is using http://example.org/person.html to identify a person, when it actually identifies an HTML document about the person. This conflation creates semantic ambiguity and breaks linked data principles.

Semantic Flow enforces clear referent distinctions through IRI patterns: slash-terminated IRIs always refer to concepts or entities, while extension-terminated IRIs always refer to retrievable content. This prevents the classic "document vs thing" confusion that has plagued semantic web implementations.

Backlinks