This work is distributed under a Creative Commons Attribution 3.0 Unported License.
This document presents VIR (VIsual Representation), an extension of CIDOC-CRM to support propositions about visual items.
Namespace: The namespace for VIR terms is http://w3id.org/vir#
Prefix: The suggested prefix for VIR is vir
RDFS: an RDFS version of the ontology is available on github
This document presents VIR, an extension of CIDOC-CRM created to sustain propositions on the nature of visual elements and permit these descriptions to be published on the Web.
With the term visual element, we refer to those signs identified in the visual space as distinct and documentable units, and subject to an analytical interpretation.
The scope of this ontology is to s to provide a framework to support the identification, annotation and interconnections between diverse visual elements and presents and assist their documentation and retrieval. Specifically, the model aims to clarify the identity and the relation of these visual signs, providing the necessary classes to characterise their constituent elements, reference, symbolic content and source of interpretation.
VIR expands on key entities and properties from CIDOC-CRM, introducing new classes and relationships responding to the visual and art historical community, specifically building up on the iconographical tradition. The result is a model which differentiates between interpretation and element identified, providing a clear distinction between denotation and signification of an element. As a consequence of such distinction, the ontology allows for the definition of diverse denotative criteria for the same representation, which could change based on traditions and perspective. Visual objects can be, in fact, polysemic and ambiguous, and it is not so easy to pin down a denotative or connotative meaning because they are very much context-dependent.
For such reasons, the model clearly separates the identity of the object and its subject matter. Moreover, while providing the means for expressing the relationships between a representation and its subject(s) (the process), the ontology does not provide any type of subject-based classification, but rely on the ones already present and widely accepted by the art historical community, such as Iconclass, the Warburg Classification or the Thesaurus Iconographique by Garnier.
The core of the ontology has been developed working with two very different projects, the documentation of the Panagia Phorbiotissa church in Cyprus and the mapping of the Berenson photo archive of the Harvard University Center for Italian Renaissance Studies. The analysis of two different visual traditions, Byzantine and Western Art, has helped the development of shared characterisation of the classes and properties which, at the time of writing, covers use cases coming from both a research and a classificatory perspective.
All the classes declared were given both a name and an identifier constructed according to the conventions used in the CIDOC CRM model.
For the classes, the identifier consists of the two letters IC followed by a number. Resulting properties were also given a name and an identifier: the letter K followed by a number.
They correspond respectively to letters "E" and "P" in the CIDOC CRM naming conventions, where "E" originally meant "entity" (although the CIDOC CRM "entities" are now consistently called "classes"), and "P" means "property".
Inverse properties share the same identifier, the letter K followed by the same number, plus the character "i" (inverse).
In this section, we present different examples in order to illustrate how VIR can be used to encode information about diverse visual items. In order to make clear the potential of the ontology, we will use diverse representations of St. George, a culturally widespread subject.
The first example is a simple one, the modelling of the statue of St George slaying the dragon in Berlin. Having already presented in the Graphical Overview the prefix used, for readability purpose we will omit this information in the next examples.
The mapping in Figure 2 presents an overview of the classification of a statue through a (IC12) Visual Recognition event. The classification is purely visual and assign a representation to an object, defining the (IC10) features and (IC16) character it portrays. The classificatory act help us track the provenance of the statements, differentiating diverse recognitions carried out by different agents, which do not share the same knowledge of the classified object.
The second example is about the information of a beautiful ceramic Flask coming from Ottoman Turkey.
The flask presents two sides, with two different representations, a standing woman and the depiction of St. George on horseback. Both representations appear in the mapping but slightly differently. For the saint, it has been used the modelling proposed in Figure 3 the statue in Berlin, but only two attributes have been recognised in this case, the horse and the spear. The other representation has been described using the property Denote, a shortcut of the full path IC12 Visual Recognition assign (K9) to a E18 Physical Thing the status of (K11) IC9 Representation. In this examples, the statue of Saint George has been classified in relation to two of its Attribute (IC10), which are modelled using a controlled vocabulary. It is important to state that the type here does not identify the attribute depicted but its iconographical type.
The third example is about the photograph of the work of art “St. George killing the dragon” by Vittore Carpaccio. The aim of this example it is not to map the information about the original work of art, or the origin of the photo, but only on the iconographical attribution and representation. The mapping is a bit more complicated than the one above, therefore it appears divided into two parts, a and b. Part a presents a mapping of the information about the attribution of the artwork.
The mapping describes two events, the recognition of the representation in the recto and, consequently, the inscription in the Verso of the photograph of the note “With (but not) Carpaccio”. The recto and verso of a photograph is described using two VIR classes which, in respect to the classic physical features, allows for clearly stating the front/back relation between the two.
The inscription is described as a (E12) production event, which results in the engraving of a set of sign in the verso, which carries a (E33) linguistic object.
The figure 5b above does complete the information in figure 5a, describing the visual information present in the photo. Several attributes were identified in this representation (dragons, princesses, lakes, vessels, castle), and two of them defined as symbolic. The dragon symbolizes the lust, while the castle the city of Silene in Libya, the place where, in the Golden Legend, St. George killed the dragon.
Representation, sometimes at least, have to be seen as the result of a long process which involves preliminary studies and sketches of what, in the end, would be the final version of an artwork. This fourth example presents the mapping of the information about a preparatory sketch and the final version of an artwork using the example of the Triumph of St. George by Vittore Carpaccio.
The mapping in figure 6 clearly present the relationship between the preparatory sketch and the final version, establishing not only a relation, K4 has visual prototype, between the two, but a relationship type. The aim of this property is to relate artwork with their derivations whatever their type. A copy is to be interpreted as using the original as its prototype, as much as an artwork use the drawing as its original form. For such reason, all the diverse type of prototypical form can be modelled using the K4.1 property and a type.
Other cases where VIR can be extremely useful would include the description of similar iconographical types which differ in attributes/features, such as episodes having the same character but diverse denotative meaning (diverse episodes of the saint life), or similar attributes over very much diverse representations (St George killing a Dragon and St. Micheal killing a dragon). The examples are numerous and the time limited, but if some help is required the issue section in the github repository of VIR would be the best place to ask.
Subclass of: E25 Man-Made Feature
An iconographical atom is a physical arrangement of forms/colours created by human activity
Subclass of: E36 Visual Item
A single pictorial item or a part of it. Single representations or region of the same representations are instance of this class.
Example:
Subclass of: E36 Visual Item
E29 Design or Procedure
A set of features considered by a viewer more salient than others and used as a key for the identification of a Representation. The attribute could correspond to iconographical elements or simple signs which the viewer uses to provide a stable identity to a visual object.
Example:
Subclass of: IC16 Character
A human, or anthropomorphic figure, that represents an abstract idea or a concept.
Example:
Subclass of: S4 Observation
The activity of assigning the iconographical status to a man-made
object, or to one of its parts. It takes into account the possibility to
link it to a speech act or a document where the authoritative
proposition is clearly made.
Example:
Subclass of: F38 Character
This class comprises fictional individuals, or groups, appearing in a
representation. Each character portrayed can have a type, for example "Saint" or "layman". Every saint portrayed is considered here as a character and not as an actor.
Example:
Subclass of: E19 Physical Feature
The front or face of a single sheet or the right-hand page of an open book. The feature is presents in object such as codex, books, pamphlets, documents, photographs and painting.
Example:
Subclass of: E19 Physical Feature
The back or underside of a single sheet of paper, or the left-hand page of an open book. The feature is presents in object such as codex, books, pamphlets, documents, photographs and painting.
Example:
Subclass of: S4 Observation
The activity of associating two or more symbolic objects (representation, text, characters, musical score, identifiable structure) using one or more dimension of resemblances between each other. The similarity encoded here is always the result of a thought process performed by an agent, whatever source he may use (his vision, the result of a computed operation, secondary sources) and it is always to be considered belonging to the domain of the visual. The encoding of other type of similarities should use other constructs. Qualities used to assess and state the resemblance are to be encoded as a E55 Type.
Example:
Domain: E18 Physical Thing
Range: E36 Visual Item
Subproperty: P65 show visual items
The property documents the assignment of an iconographical object to a specific physical man-made object. It is a shortcut for the more fully developed path IC12 Visual Recognition assign (K9) to a E18 Physical Thing the status of (K11) IC9 Representation.
Domain: IC9 Representation
Range: IC9 Representation
Subproperty: P67 refers to
The property documents the use of a specific prototypical example for an image. The nature of the relationships helps define a map of relationships between prototypical items used in the arts.
Domain: K4 is visual prototype of (has visual prototype)
Range: E55 Type
Domain: IC19 Recto
Range: IC20 Verso
Subproperty: P46 is composed of
The property documents the presence of a Verso or a Recto, respectively in the back or in the front of an object.
Domain: IC19 Recto
Range: E22 Man-Made Object
Subproperty: P46 is composed of
The property indicates the presence of a recto in the described object.
Domain: IC20 Verso
Range: E22 Man-Made Object
Subproperty: P56 bears feature
The property indicates the presence of a verso in the described object.
Domain: IC12 Visual Recognition
Range: E18 Physical Thing
Subproperty: P140 assigned attribute to (was attributed by)
The property documents the assignment of status to a specific physical
thing.
Domain: E7 Activity
Range: E89 Propositional Object
Subproperty: P16 used specific object (was used for)
The property describes the source used for the status assignment.
Domain: IC12 Visual Recognition
Range: E36 Visual Item
Subproperty: P141 assigned attribute to
The property indicates the status assigned during the status assignment
event.
Domain: E36 Visual Item
Range: E90 Symbolic Object
Subproperty: P138 Represents
The property indicates the symbolic value of the attribute presents in a representation.
Domain: E12 Production
Range: IC10 Attribute
Subproperty: P33 used specific technique
The property indicates the specific attribute used during the production of a visual object
Domain: IC9 Representation
Range: IC10 Attribute
Subproperty: P106 is composed of (forms part of)
This property associates an attribute with the representation where it is depicted.
Domain: IC9 Representation
Range: IC9 Representation
Subproperty: P148 has component (is component of)
This property put in relation a representation with a part of itself.
Domain: IC10 Attribute
Range: E55 Type
Subproperty: P137 exemplifies (is exemplified by)
This property indicates the type of object depicted by an iconographical attribute.
Domain: IC9 Representation
Range: IC11 Personification
Subproperty: P138 Represents
This property indicates the membership of a personification in an iconographical object.
Domain: IC9 Representation
Range: IC9 Representation
Subproperty: P138 Represents
This property indicates the connotation relationships, formalized by Barthes, between a conceptual entity and an iconographical object. It is a shortcut for the more fully developed path IC12 Visual Recognition assign (K9) to a IC9 Representation a new (K11) IC9 Representation. It doesn't offer any information about when and whom established the connotation relationship.
Domain: IC9 Representation
Range: IC16 Character
Subproperty: P138 Represents
This property put in relation an iconographical object with the portrayed character.
Domain: IC11 Personification
Range: E90 Symbolic Object
Subproperty: P138 Represents
This property put in relation a symbolic object with a personification in a work of art.
Domain: IC16 Character
Range: E39 Actor
Shortcut: E57 is based on
This property associates an instance of IC16 Character with an instance of E39 Actor that the character is motivated by or is intended to represent.
Domain: IC9 Representation
Range: E73 Information Object
Subproperty: P67 Refers to
This property associates an information object to a iconographical representation.
Domain: IC21 Similarity Statement
Range: E90 Symbolic Object
Subproperty: P140 Assigned Attribute To
The property document the assignment of the status of visual similarity to a symbolic object.
Domain: E90 Symbolic Object
Range: E90 Symbolic Object
Subproperty: P65 Shows Visual Item
The property documents the assignment of Similarity status of a symbolic object. It is a shortcut for the more fully developed path IC21 Similarity Statement assign similarity (K35) to two E90 Symbolic Objects.
VIR is licensed under a Creative Commons Attribuzione 3.0 Unported License.
