The scope note of E59 Primitive Value (up until v7.4) states that its instances are not considered elements of the universe of discourse the CIDOC CRM aims to define and analyse. This formulation is potentially misleading, since certain E59 Primitive Value subclasses, namely E94 Space Primitive, E61 Time Primitive, and E95 Spacetime Primitive, are explicitly declared as subclasses of E41 Appellation and are therefore characterised ontologically as names within the CRM.
This issue is about redrafting the definition of E59.
Post by Eleni Tsouloucha (22 September 2026)
Dear all,
please take a moment to review the proposed scope note edits that Martin, George and Gerald have proposed for E59 Primitive Value. The motivation is that the current scope note of E59 Primitive Value states that its instances are not considered elements of the universe of discourse the CIDOC CRM aims to define and analyse. This formulation is potentially misleading, since certain E59 Primitive Value subclasses, namely E94 Space Primitive, E61 Time Primitive, and E95 Spacetime Primitive, are explicitly declared as subclasses of E41 Appellation and are therefore characterised ontologically as names within the CRM.
The HW for this issue can be found at the following link: https://docs.google.com/document/d/1m1dRqk390UedbM0UORzIP5jQfgeVWNblkwf…
Best,
Post by Chris Banks (22 September 2026)
Dear all,
I have read the redrafted scope note. I have no view on the ontology and will not pretend to one; I write with two questions from the practitioner side, and both come out of issue 723 rather than out of this one.
The first is about the last sentence. It says that how these primitives "are or should be formulated and defined... remains outside of the purview of the ontology and the work of the domain specialist for that form of information encoding." Read narrowly, that is about encoding a date or a coordinate, and I have no difficulty with it. Read broadly, it could be taken to place any domain specialist's vocabulary for handling such values outside the purview too, and 723 asks the SIG for precisely that: a recommended minimal vocabulary for typing a documentation act. I am not suggesting the drafters intended the broad reading. I am asking whether the sentence should be narrowed so that it cannot be taken that way, because if it can, the two issues will collide at the meeting and it would be better to have noticed beforehand.
The second is smaller and is about what my records actually contain. Across fourteen analyzed records I hold fifty-five entries documenting that a property was sought and no value obtained. The properties sought fall into three groups: a party, most often P23 transferred title from; a time, P4 has time-span; and a consideration, P179 had sales price. Under the proposed text those three now have three different ontological statuses. The party is an E39 and is plainly in the universe of discourse. The time resolves to an E61 Time Primitive, which the new sentence declares a name with the qualities and inferences that follow. The price resolves through E97 to an E60 Number, which remains outside. My question is whether an E13 documenting that no value was assigned behaves the same in all three cases, or whether the act differs depending on what would have been there. I do not know the answer and it affects how I would write the next record.
Both questions are genuinely questions. If the answer to either is that I have misread the text, that is a useful answer and I would rather have it before Nuremberg than in the room.
Christopher Banks
Post by George Bruseker (23 September 2026)
Dear Chris,
The Primitive Value classes are a stand in for where the ontology really does end.
It means to say we do not define data types. The ontology defines the ontological structure enabling the description of real world facts within its domain (cultural heritage broadly) as a semantic network.
As long as those facts are representations of things within the domain, it is within our purview and mandate to represent them: a class for persons, a class for acquisition events and so on.
These primitive value classes are where the actual data value would be a data point. This is not an ontological question anymore, it is a question of how do you document a data type.
For this there are competent committees and domain specialist groups.
E.g.: OGC (Open Geospatial Consortium) for example defines WKT / WKB for geopoints
W3C defines the XSD date/time/duration types
What we are saying is we do not want nor do we have the competence to draft data types for information in information systems. When the ontology reaches out to a data type, we assume another competent committee will make one that we can adopt.
So these are the primitive values to which we refer.
I believe your issue refers to whether there could / should be a recommended vocabulary of types relative to E13 and certain domain practice. Types in CIDOC CRM relate to the class E55 Type and because it is a class, it falls within our purview in principle.
While in general CIDOC CRM is not meant to be prescriptive, we have recently discussed having some suggested vocabulary terms in some places as a guide. Initiatives like Linked. Art or SRDM which define application profiles using CIDOC CRM take stronger stands where, for a particular purpose, they document how they expect CIDOC CRM to be used to describe a phenomenon including required vocabulary.
So I think part of the meat of the issue that you raise will be, is this a phenomenon so general in practice that we should prescribe it at the ontology level or is it simply that we indicate how a general pattern from CRM can connecto a particular documentation practice that you define and find useful and woudl recommend to others (an application profile).
Cheers,
George