From W3C Wiki
Jump to: navigation, search

Semiotics of RDF "Signs"


This Wiki is an attempt to define terminology (with a companion OWL ontology) for RDF URIs as a guide for the language used when discussing issues surrounding httpRange-14 (information resources, non-information resources, dereference, etc.). The content here is heavily influenced by DanConnolly's prior work in this regard:

The goal is to draw direct correlation(s) between semiotic terminology and web architecture terminology to see what breaks (if anything).

Naming Conventions

When used here, the idRes: prefix is associated with the BaseURI: http://www.w3.org/2001/tag/fdesc54/webarch# . Some of the terms in that RDF/OWL ontology are re-used here but with a different base URI (the URL of the Architecture of the World Wide Web document) and with the 'normative' fragment identifiers in that document. So, idRes:URI becomes web:URI and idRes:Resource becomes web:def-resource (fully qualified: http://www.w3.org/TR/webarch/#def-resource). The web: prefix is associated with this base URI.

We will using the prefix sign: as a shorthand for the baseURI common to all the terms in the companion ontology: "tag:chimezie_at_metacognition.info,2007:RDFSemiotics#"

Semiotics 101

In addition, we will use, as primary input, John Sowa's summary [1] of Charles Sanders Peirce's seminal work on Semiotics:

 .. Peirce viewed semiotics or as he spelled it, semeiotic, as the unifying theme that relates naturally occurring signs and conventional symbols to concepts and reality: language, logic, and science are systems for manipulating signs to represent relationships in and about the world. In classifying signs, Peirce used his trichotomy to subdivide the sign relation into the triad of icon, index and symbol: an icon represents an object by its inherent form, which resembles the intended object; an index represents its object by some causal relationship, such as a pointing finger, a weather vane, or a dial on a meter; and a symbol represents its object by a convention established by some community.


The following terms are re-used from that latter body of work to reduce ambiguity:

  • Sign - something which stands to somebody for something in some respect or capacity.
  • Referent - the 'thing' denoted by a sign
  • Interpretant - a sign created in the mind of an agent / consumer of signs which is either equivalent or more developed ("Concepts)
  • Icon - a sign which represents an object by its inherent form
  • Index - a sign which represents its object by some causal relationship, such as a pointing finger, a weather vane, or a dial on a meter
  • Symbol - a sign which represents its object by a convention established by some community
  • denotes - the relation which holds between signs and their referents

Semiotic Triangle


Chapter 2 ("2. Signs of Signs") of Ontology, Metadata, and Semiotics [2] has a very useful diagram which describes the relationships between referents, signs, and interpretants. A decent analogy is included there:

 The terms morning star and evening star are distinct signs that create different concepts or interpretants in the mind of the listener. Both concepts stand for the same object, but in respect to a different ground, which depends on the time of the observation.

Semiotic Upper Ontology

Below is an 'upper' OWL ontology (the classes are expressed in the Manchester OWL syntax [3] - for brevity - and the properties are expressed in OWL's abstract syntax) which captures the exhaustive categories of signs:


Class: sign:Sign
DisjointWith: sign:Referent

Class: sign:Referent
DisjointWith: sign:Sign

Class: sign:Interpretant
SubClassOf: sign:Sign
DisjointWith: sign:Icon or sign:Index or sign:Symbol

Class: sign:Icon
SubClassOf: sign:Sign
DisjointWith: sign:Referent or sign:Index or sign:Symbol

Class: sign:Symbol
SubClassOf: sign:Sign
DisjointWith: sign:Referent or sign:Index or sign:Interpretant

Semiotic Convention for Web Architecture

Finally, below are OWL expressions which place the Web Architecture notions of information resource, HTTP dereference, and non-information resources in their proper location within the above ontology. The web2: prefix is used without a baseURI and is associated with terms not explicitly defined in AWWW but are determined via interpretation of the relevant informative language. Finally the http: prefix is bound to the URI "http://www.w3.org/2006/http#"





Class: web:def-representation

Class: http:Response
SubClassOf: web:def-representation

Class: web:URI
SubClassOf: sign:Symbol that ( idRef:identifies some web:def-resource )

Class: web:def-resource
EquivalentTo: sign:Referent

Class: web:def-information-resource
EquivalentTo: web:def-resource that ( idRef:representation only ( http:Response that http:responseCode only ( http:200 or http:201 .. or .. http:206 ) ) )

So, for example, members of the web:def-information-resource class are precisely those resources which meet the criteria for information resources as defined by the AWWW document (specifically the section identified by the representation returned when the URI is dereferenced).

Semiotic Classes of URIs

Finally, we introduce three classes of URIs (signs): those which denote (or identify) information resources, those which denote "non-information-resources", and those which don't resolve to anything (pragmatically, those URIs in a scheme without a resolution mechanism or which respond with a 4XX to a HTTP GET request). The sign: prefix is used again for these terms:

Class: sign:IRIdentifyingURI
EquivalentTo: web:URI that ( idRef:identifies only web:def-information-resource )

Class: sign:NonIRIdentifyingURI
EquivalentTo: web:URI that not sign:IRIdentifyingURI

Class: sign:NonResolvableURI
EquivalentTo: sign:NonIRIdentifyingURI and not ( web:dereference-uri some web:def-representation )

So, the URI used here to denote the class of information resources is itself a member of the sign:IRIdentifyingURI class


Some notes: According to the httpRange-14 resolution, a 2XX response (to an HTTP GET: web2:http-get-uri) indicates that the result of dereferencing the URI is a representation which captures the state of the very web:def-information-resource denoted by this URI. In this scenario the idRef:indentifies relation is sufficient since, what is identified is also directly described by the representation which was 'resolved'. However, in the case of a 303, the representation returned (via web2:http-get-uri) does not describe the original resource (referent). Quoting Roy Fielding's email regarding the resolution and 303:

  ... a mechanism whereby information can be supplied via the 303 redirect without leading to ambiguous interpretation of such information as being a representation of the resource (rather, the redirection points to a different resource in the same way as an external link from one resource to the other).


  1. http://www.jfsowa.com/pubs/signproc.htm
  2. http://users.bestweb.net/~sowa/peirce/ontometa.htm
  3. http://owl-workshop.man.ac.uk/acceptedLong/submission_9.pdf