GregorHagedorn - Fri Nov 19 2004 - Version 1.2 Parent topic: LCProtonymDiscussion About cases of botanical combination and nomen novum (replacement names):
SCENARIO 1
Gregor: A validly published name Genus1 xyz Author1 is combined into a new Genus 2 Author2 in which the epithet xyz is already used by a different species, a nomen novum is created: Genus2 zzz. Now someone else has a different opinion, and combines the species into Genus3. Would that be Genus3 xyz or Genus3 zzz?
Paul Kirk: Genus1 xyz Author1 (year 1) has the epithet preoccupied in Genus 2 so Genus2 zzz Author2 (year 1+) is introduced as a nom. nov. Genus1 xyz Author1 = Genus2 zzz Author2 (homotypic synonyms) are to be placed in Genus3. If epithet xyz is not preoccupied in Genus3 the correct name is a combination of the correct genus and the final epithet of the legitimate name which has priority at the intended rank - hence Genus3 xyz (Author1) Author3. There is no choice unless conservation is invoked for example where epithet zzz is well known. ICBN article 11.4 is the relevant Article.
SCENARIO 2
Gregor: Genus1 xyz Author2 is a later homonym of Genus1 xyz Author1. A nomen novum Genus1 zzz Author3 is a replacement name for Genus1 xyz Author2. Someone else has a different opinion, and combines the species into Genus3.
Paul Kirk: Genus1 xyz Author2 (year 1+) is illegitimate because of Genus1 xyz Author1 (year 1). Genus1 zzz Author3 (year 1++) is the legitimate (assumed valid) nom. nov. The correct name for this species is the correct generic name (Genus3) and the final epithet of the legitimate name which has priority at the intended rank - hence Genus3 zzz (Author3) Author 4 - Again ICBN Article 11.4. is the relevant Article. -- The above assumes no additional heterotypic names are involved in the circumscription (synonymy). Art. 58.1 is there to deal with illegitimate names (something Zoologists do not suffer from) and impose some conformity.
Conclusion for LC schema
Both "comb. nov." and "nom. nov." can be based on either "original name" or "replacement name", depending on the situation. Thus the relation-types "basionym" and "replaced synonym" point to a protonym and not to a more specific name record type.
A comment from Jerry's original LC schema proposal that I would like to preserve for the moment (it is gone in the newer schema versions now): "In the case of ICBN there are some subtle rules for using nom. novs. 1 the original name is illegitimate in which case the nom. nov. has itself as a basionym and the replaced name is the illegitimate name. 2 a combination would result in a homonym or tautonym, in which case the nom. nov. has itself as a basionym and the replaced name is the original valid name. 3 a combination would result in a homonym or tautonym and the author selects the next available epithet in the original circumscription. In this case the basionym is that later name (even though it is heterotypic!) and the replaced name is the original name. NB In this rare case linking names on the basionym INCLUDES AMBIGUOUS SYNONYMS."
Gregor: I would like to be able to distinguish between base-name relation not known, and not present. "Original names" can not points to another name, but they would be identical to combinations or replacement names where the link is not (yet) known. Two solutions come to my mind: Either derive knowledge from typing the names (having a type "original name", compare LCNameCreationTypeDiscussion), or by setting proto-/basionym pointer to own ID for those records known to be an original name that are not based on a different name. (I think above Jerry assumes the latter)