GregorHagedorn Thu 03-06-2004 14:25

Why has the name no publication? How am I supposed to find the original publication of the concept name? I guess most of the information that you would expect is delegated to a generalized relation method, whose type enumeration seems to be very short on the things the nomenclatural knowledge wants to express. I particularly miss (without doing detailed checking; just simulating a taxonomist trying to expert her or his data to the standard): Lectotypification (this acutally redefines the concept of the name), designation of epitype, validation, nomen novum, recombination, sanctioning (only fungi), anamorph/teleomorph relation (fungal terms, but same happens in animals with larval names). All this is from a botanical perspective.

You would have to provide quite extensive documentation to document how to make this expressable and retrievable using the relation.

Some of these are appropriately dealed with a concept relations, esp. lectotypification and designation of epitype. These are a bit a mix, since priority rules apply, but in essence they provide replacement concepts where the original concept is no longer operational. The fact that they have a higher nomenclatural status does not bother me that much. Similarly, it is fine with me to treat the original concept as a concept.

Other relations I think are much easier if you treat them as part of the name, as it is done in botanical usage.

"Genus species Author1 ex Author2" (validation) is identical with "Genus species Author2" (the first publication had to be validated, making the latter the true author), whereas

"Genus species Author1 in Author2" is identical with "Genus species Author1" (the publishing source may be added as a convenience, but can be omitted).

I fear that retrieving such knowledge from the general relations may be very complicated.

RobertKukla Thu 03-06-2004 16:09

I am correct in assuming your examples above are 4 different name entries in your proposed schema?

GregorHagedorn Thu 03-06-2004 17:29 No, with nomenclatural knowledge you know the identity of the two pairs. You have 4 namevariant strings, but only two names. I have to know about the two publications in the ex case (which would be two different concepts in Napier with a relationship between them, however). Nothing makes the figuring out impossible in the relation model, my point is am missing the ability to express the identity if you don't provide name identifiers. I am expressing a relationship between names here, not concepts.

If 100 publications circumscribe "Genus species Author1 ex Author2", each can be cited using both name variants.

I believe in the Napier model I would have to express that knowledge for each of the 100 concepts, rather than expressing it once.

RobertKukla Thu 03-06-2004 16:09

In most data sources I have seen "in" and "ex" authors are part of the name string and not parsed out in to separate fields. If such reasoning is required, why could it not be done on the fields of the name of the original concept?

GregorHagedorn Thu 03-06-2004 17:29

I agree and I use a parser to do this. Some database like the DiversityWorkbench?? separates these author parts out, others don't.

However, I would like to store that knowledge that I have figured out the correct name object. That would allow me to go directly to GenBank?? based on a shared name id, rather than downloading ALL their names (which is impossible in fact since not supported...), then parsing/comparing them to find the answer.


let's make this a use case: How does the Napier model propose to interact with Genbank?