On 9 December 2010 22:31, Richard Pyle deepreef@bishopmuseum.org wrote:
I guess what I don't understand is: what would go in scientificNameWithAuthorship that isn't already achievable via these three terms:
verbatimScientificName (any text string purporting to represent as unambiguously as possible a scientific name, inclusive of authorship if available) scientificName (effectively canonical name explicitly without authorship = your scientificNameWithoutAuthorship) scientificNameAuthorship (only the authorship; no scientific name elements)
Give me a use case where one would want to use something like scientificNameWithAuthorship in a way that couldn't be served by these three elements.
Use case: A database desiring to identify taxa unambiguous by nomenclatural name would have a field containing:
Lobelia spicata Lam. var. spicata
I can export this to verbatimScientificName, but I cannot inform anyone that verbatimScientificName contain a full canonical name with authorship. As a consumer, I must expect verbatimScientificName to contain anything like: Lobelia spicata Lam. var. spicata Lobelia spicata var. spicata L. spicata Lam. var. spicata Lobelia sp. var. spicata (a common form!) Lobelia (Lobelia) spicata Lam. var. spicata (Lobeliaceae) or, in the case of a non-autonym, the canonical form would be: Lobelia spicata var. campanulata McVaugh but many many databases may have: Lobelia spicata Lam. var. campanulata McVaugh
What I am striving for is a field in which the best canonical form available to the provider can be expressed. If a provider has canonical form, this can be expressed as: verbatimScientificName = Lobelia spicata Lam. var. spicata scientificNameWithAuthorship = Lobelia spicata Lam. var. spicata
if the provider knows that it cannot provide a canonical name, then as: verbatimScientificName = Lobelia spicata Lam. var. spicata scientificNameWithAuthorship =
Note that this also applies to zoological databases that did not split author and name. It is impossible to provide a name that is known to be a canonical name including the authorship to the present proposal. You can put that into verbatim, but the consumer may not have any quality expectation on the verbatim name.
*plenty* of rare use cases out there that DwC does not accommodate, so I don't see that as justification for introduction of a new term, that might further confuse people. Moreover, it seems like a very simple algorithm for a consumer to recognize an autonym (nomenclaturalCode=ICBN + Rank is below species + second two components of trinomial are identical), and then format the string accordingly from scientificName and scientificNameAuthorship.
I doubt this. You would have to be able to correctly parse all names in all variants, including non-canonical and mistreated ones.
My point is: with the present proposal you require to parse ANY name... You cannot have any expectations on a fully qualified name including authorship with the present proposal. With the modified proposal, you can make quality expectations on most names, and limit any extra work to those that have no canonical form.
I *STRONGLY* disagree with your suggestion to drop scientificNameAuthorship. This is an extremely fundamental component to nomenclatural disambiguation, and a relative "pain in the parse" for a consumer when provided only with a full name-string-with-authorship. To me, the use cases where your suggested scientificNameWithAuthorship cannot be easily met with a combination of verbatimScientificName , scientificName, and scientificNameAuthorship are far, far, far fewer than the use cases that would benefit from receiving content where authorship is pre-parsed from canonical name.
I think we must have a misunderstanding here: The most common way to express a taxon name is with authorship. This does not solve all problems of concepts, but it does solve the problem of homonyms. Most journals require this (let us not debate whether instead they should require tdwg-lsids). Most online systems show names with authors. Why then is the desire to have this form where the downstream user of dwc data can have an assurance of this form of name a rare use case?
Gregor