Add function PROPERTY as a synonym for PREDICATE

[afs]

"property" and "predicate" get used interchangeably.

Should we have a synonym PROPERTY for PREDICATE?
(The grammar currently has PREDICATE)

c.f. isURI/isIRI and URI/IRI

[domel]

I think this question can be asked more broadly. There have already been a few reports that I can classify as minor (those that do not break backward compatibility with SPARQL 1.1). One example was the proposal to add SHA3 functions. I think WG should decide where the boundary is for adding new functions to SPARQL 1.2 and what should be reported to sparql-dev as SEP.

[afs]

We need to consider the SHA3 functions - it something that can be done in "new features" mode. There are a lot more hash functions these days.

This issue is a little different though - it's for a synonym of an existing function, not adding a function.

PREDICATE is already in the spec as a triple term accessor and is necessary for RDF 1.2/SPARQL 1.2 - the spec already has a precedent for synonyms with isURI/isIRI.

[hartig]

I don't see a need for adding such a synonym for this function name (but I will also not fight against it if many here would like to see this added).

Having both isURI and isIRI has historical reasons I assume.

[rubensworks]

-1 from me as well. Offering multiple ways to achieve the same thing often adds more confusion than benefit IMO.

[afs]

URI and IRI were already being used interchangeably even back in SPARQL 1.0 days. The HTTP URL spec is UTF-8.

We all use predicate/property interchangeably in common usage so, as I see it, it is already "out there".
(I can't think of anything in this category.)

Given an open choice - which of PREDICATE or PROPERTY do you prefer?

[hartig]

I would prefer PREDICATE over PROPERTY.

[kasei]

I'd be in favor of supporting both names. As @afs says, they are often used interchangeably. With the potential of rdf-star to bring more visibility and potential use from LPG communities, I think PROPERTY is also a more inclusive term (but not to the exclusion of PREDICATE which obviously has deep history in the RDF/SPARQL community).

I think the existence of isURI/isIRI has shown that there hasn't been much (if any) confusion stemming from the availability of synonyms.

[domel]

While I understand the argument for supporting both terms, I would caution against using PROPERTY in this context, particularly given its specific and different connotation in LPG communities. In LPG models, PROPERTY refers to key-value pairs associated with nodes or edges, not to relationships or predicates themselves. This creates a risk of confusion when engaging with those communities, especially as RDF-star seeks to bridge the gap between RDF and LPG paradigms.

Predicate, on the other hand, has a long-standing and precise definition in RDF and SPARQL, ensuring clarity within these communities. Retaining "Predicate" as the primary term would help avoid ambiguity while respecting the established terminology and semantics in the RDF ecosystem.

[niklasl]

I also prefer PREDICATE over PROPERTY. Given a wide enough audience, would there be a shared expectation of what PROPERTY(<<(skos:prefLabel rdfs:subPropertyOf rdfs:label)>>) should return (given that each IRI denotes a property)?