search

tdwg / tcs2

The TCS 2 Task Group will turn TCS into a form in which it can be maintained. The new version of TCS will be a vocabulary standard like Darwin Core and Audiovisual Core and will complement these other existing TDWG standards.
6 stars 0 forks source link

property:traditionalSynonymyRelationshipType #205

Open nielsklazenga opened 1 year ago

nielsklazenga commented 1 year ago

traditionalSynonymyRelationshipType (property)
Label Traditional synonymy relationship type
Definition Type of relationship that is used in traditional synonymy, e.g. 'pro parte synonym' or 'misapplication'.
Usage notes
Comments This term should be used together with relationshipType (which is required anyway). Generally, the value of relationshipType will be intersects if this property is used.
Required No
Repeatable No
Constraints Controlled term
nielsklazenga commented 1 year ago

This term needs a champion (or champions) to provide comments and the controlled vocabulary if we are going to have it. Personally, I would rather (have people) use comments (rdfs:comment) for this.

nielsklazenga commented 1 year ago

@camwebb in PR #206:

...so I can't remember the discussions about this that clearly. The relevant GH discussion seems to be in #65, in particular a potential disagreement (between @Archilegt and you) about whether synonyms should be properties of names or TCs. I agree with you that to make a synonymy statement one must have a TC to work with, and in general the most accurate way to encode synonymy is using TC intersects TC. But... in practice most synonymy statements have no "sec." for the names so they appear to be properties of names. Have we discussed writing any general guidance for users which would allow/encourage TC name X; accordingTo NULL? I.e., forcing a name to be a concept, but with unknown TaxonTreatment. If we will allow this, then it makes sense to model synonym as a TC property. If this is not allowed, we must make synonymy a property of names, otherwise vast amounts of taxonomy information cannot be correctly encoded.

nielsklazenga commented 1 year ago

This is not about synonymy, but about taxon relationship statements, like "pro parte synonym" (#200) and "misapplication" (#199) masquerading as synonymy. These should be treated as intersects taxon relationships, if they are treated at all. This property is just to qualify the intersects relationship type. I'd rather people do that in remarks, if they really have to, but if people insist on having something they can put a controlled vocabulary on, I think this is fairly innocuous. Having the controlled vocabulary, by the way, is a condition for having this property, so if people want this, they have to do the work. I note that controlled terms also need definitions.

Having synonym (#65) as a property on the Taxon Concept – which we have had for two years now – makes accepted name (taxonName #2) and synonym (#65) equivalent(ish) to skosxl:prefLabel and skosxl:hiddenLabel, respectively, and that is how it should be. Synonymy statements should not have an accordingTo: they are not assertions. That two specimens that happen to be the types of names belong to the same taxon is in the Taxon Concept (SpecimenCircumscription in TCS 1) and the rest is applying nomenclatural business rules.

Synonymy – being part of nomenclature – is about applying labels to things, not about relationships between things. It is important that we stop confuddling the two. So, while I understand (or am trying really hard to) that people want to find a place for terms like "pro parte synonym" and "misapplication", it would really be best to remove terms like these from our discourse altogether.

If it is not already clear, I am not in favour of having this term, but people can have it if they can land it.